Enterprise AI startup RapidCanvas has raised $29 million in a Series B funding round as it looks to expand its AI engineering teams in India, the US, and Brazil.

The round was led by Vertex Ventures SEA & India. Hitachi Ventures joined as a new investor, while Accel, Peak XV Partners, Valley Capital Partners, and Titanium Ventures participated again reported Dealstreetasia.

The funding is a sign that investors are looking beyond AI chatbots and demos. They are backing companies that can help businesses turn AI ideas into working tools inside factories, finance teams, supply chains, and other daily operations.

What RapidCanvas does

RapidCanvas was founded in 2021 by Rahul Pangam and Uttam Phalnikar. Pangam is the company’s co-founder and CEO, while Phalnikar is co-founder and CTO.

The company works in an AI execution business. That means it does more than sell software and leave customers to figure it out. RapidCanvas combines its AI platform with engineers who work with customers to understand a business problem, build a custom solution, and help run it after launch.

For example, a manufacturer may have large amounts of data about inventory, machine output, supplier delays, and customer demand. RapidCanvas could help build an AI tool that identifies patterns, predicts delays, or helps production teams make faster decisions.

The company’s model is built around the idea that AI projects often fail when they remain stuck at the testing stage. Many businesses can experiment with AI, but putting it into real workflows safely and reliably is harder.

Why India is important to the expansion

As per Mint, RapidCanvas already has an office in Hyderabad and is now planning to grow its field AI engineering team across India, the US, and Brazil.

India offers a large base of software engineers, data professionals, and domain experts. For RapidCanvas, this could help it build and deploy AI applications for global customers while also serving Indian enterprises that are beginning to invest more actively in AI.

The company will also use the new capital to expand its go-to-market teams and its network of private-equity and advisory partners. In simple terms, it wants more people who can sell, support, and deliver its AI products to businesses.

A different approach from traditional IT services

RapidCanvas is entering a market traditionally served by large IT consulting firms and data-science teams. Instead of relying only on long consulting projects, it combines AI agents with engineers who remain involved in delivering results.

The company says its field AI engineers learn how a customer’s business works, create tailored applications, and remain accountable after the tool goes live.

This puts RapidCanvas in competition with enterprise AI platforms such as Databricks, Altair, and Quantexa, as well as large IT services companies building their own AI offerings. The competition will depend on which companies can deliver useful results without creating expensive, complicated systems.

What the $29M funding means

The Series B takes RapidCanvas’s total funding to $52.5 million, according to the company’s latest funding details.

The money gives the startup more room to hire engineers, improve its platform, and win enterprise customers. But the bigger challenge is proving that its AI tools can create measurable value. Businesses are becoming more careful about AI spending and want clear gains such as lower costs, faster operations, better forecasting, or improved customer service.

RapidCanvas has positioned itself around this gap between AI experimentation and real deployment. Its growth in India will be watched closely because the country is both a major engineering talent centre and an increasingly important market for enterprise AI.

Hyundai Motor Group is making one of its biggest long-term bets on India. The South Korean auto group has announced a $15.5 billion investment plan for the country between 2026 and 2032.

The headline figure is large, but one detail stands out. Hyundai plans to direct around 20 percent of the investment, or roughly $3.1 billion, toward new businesses that include artificial intelligence, venture investments, and startups.

This signals that Hyundai sees India as more than a market for selling cars. It wants the country to become a manufacturing, engineering, technology, and mobility base for the group.

How Hyundai plans to use the $15.5B investment

Hyundai says the seven-year investment will cover manufacturing expansion, research and development, supplier development, localized parts production, and future technology.

According to Hyundai Motor Group Vice Chairman Jang Jae-hoon, about 60 percent of the total investment will be used to upgrade existing facilities and expand production capacity. That works out to approximately $9.3 billion.

As per Yonhap News, around 20 percent, or about $3.1 billion, is planned for new business areas such as AI, venture investment, and startups. The remaining share is expected to support research and development and related areas as per the reports.

Hyundai also plans to establish a dedicated investment entity in India, called HMG India, to coordinate these investments.

Why AI and startups matter to Hyundai

Cars are changing quickly. Modern vehicles are no longer only mechanical products. They increasingly depend on software, connected services, digital displays, battery systems, advanced safety features, and data.

AI can help Hyundai in many ways. It can support factory automation, predict maintenance needs, improve quality checks, analyze supply-chain risks, and create smarter in-car features. In the future, AI could also play a bigger role in driver-assistance systems, connected vehicles, mobility platforms, and charging networks.

Startup partnerships may help Hyundai move faster in these areas. Instead of developing every technology internally, the company can work with young firms building solutions in electric mobility, battery technology, vehicle software, AI tools, recycling, charging, logistics, and hydrogen.

Hyundai had already launched its Innovation Challenge 2026 in India to work with startups in fields such as electrification, connectivity, advanced driver-assistance systems, AI-led tools, and next-generation vehicle technology.

A bigger push for local manufacturing

The majority of Hyundai’s proposed investment will still go into its core automobile business. That includes factory upgrades, higher production capacity, local supplier development, and vehicle manufacturing.

This is important because India is both a major vehicle market and a possible export base. Localizing more parts can reduce dependence on imports, improve supply-chain control, and make vehicles more competitive on price.

Hyundai has also said it will explore opportunities in EV charging, automotive financing, end-of-life vehicle recycling, and other mobility-linked services. These areas could create new business opportunities beyond car sales.

Competition will shape the strategy

Hyundai’s India plan comes as competition becomes stronger. Maruti Suzuki remains a major force in passenger vehicles, while Tata Motors has built a strong presence in electric vehicles. Mahindra is expanding its EV ambitions, and global brands such as Kia, Toyota, MG, and BMW are also increasing their technology focus.

For Hyundai, investing only in factories may not be enough. It needs stronger local technology partnerships, faster software development, competitive EV products, and a reliable charging ecosystem.

The AI and startup allocation does not guarantee success. Hyundai will need to choose useful technologies, support promising founders, and turn pilots into products that customers actually want.

Image

Google is widening its AI efforts in India beyond consumer tools and chatbots. Its latest announcements focus on three very different groups – small businesses, agricultural developers, and teachers.

The common idea is practical use. Google wants AI to help an MSME find credit, allow a developer to build a farm advisory tool, and give teachers more support with lesson preparation. The success of these programmes will depend on whether they work well outside big cities and for people with limited technical knowledge.

Google.org commits $15M for MSME AI tools

Google.org has announced $15 million in funding for The/Nudge to support AI-led tools for India’s micro, small, and medium enterprises, commonly called MSMEs.

The proposed programme will focus on familiar problems for small business owners – access to formal credit, compliance work, and finding new markets. Many small firms still manage accounts manually, struggle with tax or licensing processes, and find it difficult to prove their creditworthiness to lenders.

The funding is expected to support an open call for solutions, where teams can develop tools for these challenges. It also includes plans for a multilingual, voice-led skills platform aimed at people in Tier 3 and Tier 4 towns.

A practical example could be an AI tool that helps a small manufacturer organize invoices, understand compliance deadlines, and prepare clearer business records for a bank loan. But the real test will be accuracy, language support, affordability, and whether small businesses trust the tools with their data.

The money is a philanthropic grant, not an investment in a startup as per ET.

Agricultural AI APIs open more options for developers

Google has also expanded access to its agricultural AI models, allowing developers, startups, researchers, and public-sector teams to build local farming tools.

The two key tools are the Agricultural Landscape Understanding API and the Agricultural Monitoring and Event Detection API. In plain language, they use satellite data and machine learning to help identify farm boundaries, water bodies, crop patterns, sowing periods, and harvest activity.

This could help create useful services for farmers. A local app, for instance, could combine crop information with weather data and send timely alerts about irrigation, sowing, or possible risks. Governments and agri-tech companies may also use the data for planning, insurance, procurement, and resource management.

Google says its models were first built for India and are now being made available more widely across Asia-Pacific. The company cited Gujarat’s Krushi Pragati platform, where the APIs have been used to support large-scale SMS-based farm advisories as per Google India Blog.

However, AI-based farm advice should support human decision-making, not replace it. Crop conditions can change quickly because of weather, pests, soil quality, and local practices. Farmers need information that is local, clear, and reliable.

More than 80,000 educators trained in AI

Google says it has trained more than 80,000 educators in India through its Google AI Educator Series since the programme began in May 2026.

The programme is now expanding through partnerships with Bihar, Madhya Pradesh, Sikkim, and the Ministry of Skill Development and Entrepreneurship. Google is also making the training available as a free, self-paced online course.

The training is designed to help teachers use AI responsibly for tasks such as drafting lesson plans, creating practice material, and organizing routine work. It is aligned with the National Education Policy 2020. The course is currently available in English, while versions in Hindi, Marathi, Telugu, Assamese, Odia, and Punjabi are planned.

For teachers, AI can save preparation time. But it cannot replace subject knowledge, classroom judgment, or the human connection that students need. Educators must still check AI-generated information for errors, bias, and age suitability.

A wider race for India’s AI ecosystem

Google is not alone in seeing India as a major AI market. Microsoft, Amazon Web Services, OpenAI, Indian AI startups, and cloud companies are also competing to serve businesses, developers, schools, and government organizations.

Google’s approach stands out because it links funding, developer tools, and training. That combination could help move AI from large enterprises into classrooms, farms, and smaller businesses.

Image- Google

India’s startup funding scene is moving across very different worlds this week. One young company wants to make lunar missions easier. Another is helping businesses build custom AI tools. A listed home-services platform is considering fresh capital as competition gets tougher.

The three stories involve different markets, but the aim is similar – raise enough money to build faster, hire the right people, and defend a place in a crowded future.

Cosmochute gets $1.5M for lunar mobility

Ahmedabad-based space robotics startup Cosmochute has raised $1.5 million in a pre-seed round led by Prosus Ventures. The company is part of Atoms X, a frontier-technology programme launched by Prosus Ventures with Accel.

Cosmochute was founded in 2026 by Janhavi Gore, Amit Chowdhary, and Dhruv Alok Sangal. Gore is the CEO, Chowdhary is the COO, and Sangal is the CTO. The startup is building LEAP, a modular mobility platform intended to transport and deploy payloads on the Moon.

The company’s idea is simple to understand, even if the engineering is not. Sending a machine to the Moon is difficult and expensive. Once it lands, moving instruments, supplies, or commercial payloads across the surface is another challenge. Cosmochute wants reusable rover infrastructure that could make those missions more practical for science teams and future commercial customers.

The new money is expected to support product development, engineering hiring, and space research. It is an early-stage bet on the growing lunar economy rather than a business with near-term consumer sales as per Business Standard.

RapidCanvas raises nearly $29M for enterprise AI

RapidCanvas, an enterprise AI company founded in 2021, has raised nearly $29 million in a Series B round led by Vertex Ventures and Hitachi Ventures.

The Austin-based company, led by founder and CEO Rahul Pangam, helps businesses create custom AI applications. Its platform combines AI agents with human specialists, aiming to make it easier for companies to use their own data in finance, manufacturing, supply-chain, and operational workflows.

For example, a manufacturer could use such a system to study production data, identify delays, and help staff make faster planning decisions. The goal is not simply to offer another chatbot. RapidCanvas is targeting companies that need AI tools built around their own processes and information.

RapidCanvas has not publicly detailed every planned use of the new funds. Still, the funding signals investor interest in enterprise AI platforms that help businesses move from AI experiments to real deployments. It competes in a busy market where companies must prove that their tools are reliable, secure, and useful beyond a demo as per Axios.

Urban Company considers fresh capital

Urban Company, founded in November 2014 by Abhiraj Singh Bhal, Varun Khaitan, and Raghav Chandra, is considering a fresh fundraise. Its board was scheduled to meet on October 16 to examine possible routes, including a qualified institutional placement, or QIP, and other permitted options.

A QIP allows a listed company to raise funds from qualified institutional investors. It can give a company more financial room without taking the longer route of a public offer.

The company has not confirmed the final amount or specific use of proceeds. However, reports indicate that the discussion comes as Urban Company faces stronger competition in quick home-help services from rivals such as Snabbit and Pronto. Rapid service delivery, customer discounts, worker onboarding, and marketing all require capital.

This is different from Cosmochute’s early research-focused round and RapidCanvas’s expansion funding. Urban Company is an established public company deciding how much firepower it needs to protect growth in a competitive consumer-services market as per Moneycontrol.

Every photo stored in the cloud, online payment, video call, AI prompt, streaming show, and mobile app depends on physical machines somewhere. Those machines sit inside data centers.

A data center is often described as “the cloud,” but there is nothing weightless about it. It needs land, buildings, power lines, backup systems, cooling equipment, water in some designs, and thousands of pieces of hardware. As demand for AI, digital payments, e-commerce, and connected vehicles rises, data centers are becoming an important part of modern infrastructure. They are also raising serious questions about electricity, water, and emissions.

What is a data center and why is it built

A data center is a secure facility that stores, processes, and sends digital information. It may support a bank’s mobile app, a hospital’s records, an automobile company’s connected-car platform, or an AI service.

Companies build or rent data-center space to keep services available around the clock. The facility protects equipment from outages, overheating, cyber risks, fire, and physical damage. It also helps businesses handle large numbers of users without their website or app slowing down.

Cloud companies such as Amazon Web Services, Microsoft Azure, and Google Cloud compete to offer this computing capacity. Telecom firms, banks, startups, government departments, and large manufacturers also operate their own facilities or lease space from specialist operators.

The main parts inside a data center

The servers do the actual computing. They run websites, databases, apps, AI models, and business software. Storage systems keep files, videos, backups, and customer records. Networking equipment, including switches and routers, moves data between servers and out to users.

But the equipment around the servers is just as important. A data center needs uninterrupted power supply systems, batteries, transformers, backup generators, cables, fire-suppression systems, security controls, and cooling equipment.

Land matters because a large campus needs room for the building, power substation, diesel generators, cooling systems, water infrastructure, access roads, and future expansion. Reliable electricity is usually the first deciding factor when choosing a location.

Cooling is essential because servers release heat continuously. Depending on the design, a facility may use air cooling, chilled water, cooling towers, outside air, or direct liquid cooling. Water use can be low in an air-cooled site and much higher in a facility using evaporative cooling towers.

A 1 MW data center – the basic calculation

To keep the example clear, assume the data center has a 1 MW IT load. In simple terms, its servers, storage, and networking equipment use 1 MW of electricity.

Now assume a Power Usage Effectiveness, or PUE, of 1.50. PUE measures total facility energy divided by IT-equipment energy. A PUE of 1.50 means that for every 1 kWh used by computers, another 0.5 kWh is used by cooling, power conversion, lighting, and other building systems.

The global average PUE reported by Uptime Institute was 1.56 in 2024, while newer large facilities can do better than older sites as per Uptime Institute.

With a PUE of 1.50, the 1 MW IT load becomes a 1.5 MW total facility load.

If electricity costs Rs. 8 per kWh, the energy bill alone works out to roughly:

This does not include land, construction, generators, equipment replacement, employees, network connectivity, taxes, water treatment, or maintenance.

How many servers can 1 MW support

There is no fixed answer because a basic business server may use a few hundred watts, while a powerful AI server can use several kilowatts or more.

For a simple illustration, if each compute server uses an average of 2 kW, 1 MW could power about 500 servers if all IT capacity went to those servers. In reality, some power must also go to storage and networking, so the actual number could be lower.

If those 500 servers each had two 64-core processors, the facility could contain around 64,000 physical CPU cores. That is not the same as a fixed level of performance. AI workloads depend heavily on GPUs and accelerators, while website hosting may depend more on memory, storage speed, and network capacity.

The US Department of Energy notes that server power varies widely, from mainstream systems to mid-range and high-end machines using roughly 2 kW and 12 kW respectively.

Water use and carbon emissions

Water use depends heavily on local weather and cooling technology. Using a planning estimate of 1.8 litres per kWh of total facility energy, a 1 MW IT data center with a 1.50 PUE could use about 23.65 million litres of water annually at the site. This is an estimate, not a universal figure. An air-cooled facility may use much less on-site water, while a cooling-tower design in a hot climate may use more as per US Department of Energy.

For carbon, assume the Indian grid factor of 0.716 kg CO2 per kWh, based on the Central Electricity Authority’s FY 2022-23 figure that includes renewable energy. The 1 MW example would produce about 9,408 tonnes of CO2 annually from electricity use, unless the operator matches or replaces grid power with cleaner energy.

What changes at 10 MW, 50 MW, and 100 MW

The same formula scales quickly. The table below assumes the same 1.50 PUE, Rs. 8 per kWh electricity cost, 1.8 litres per kWh water estimate, and 0.716 kg CO2 per kWh grid factor.

Larger data centers may achieve better PUE because they can invest in efficient cooling and power systems. Still, better efficiency does not automatically mean lower total consumption. A huge, efficient facility can use more electricity overall than a smaller, less efficient one.

Are data centers a threat to the environment

They can be, especially when rapid construction depends on coal-heavy electricity, scarce water, or poorly planned land use.

Electricity is the biggest concern. The International Energy Agency estimates that data centers used around 415 TWh globally in 2024, or about 1.5 percent of worldwide electricity demand. If that electricity comes from fossil fuels, emissions rise.

Water is another local issue. A data center may be manageable in a water-rich area but controversial in a drought-prone district. The choice between water-saving cooling and energy-saving cooling can involve difficult trade-offs.

Land use also matters. A large campus can replace open land, affect drainage, add noise from generators and cooling equipment, and disturb nearby biodiversity if site selection is weak. Then there is electronic waste. Servers, batteries, networking equipment, and cooling parts must be repaired, reused, refurbished, or recycled responsibly instead of being discarded early.

Local power supply can become strained when several large facilities seek grid connections in the same area. This can require new substations, transmission lines, and generation capacity.

Why data centers still matter

Data centers support digital payments, emergency services, research, cloud software, online education, connected vehicles, manufacturing systems, and startups that cannot afford to build their own IT infrastructure. The answer is not to stop building them. It is to build and operate them more carefully.

Renewable power purchase agreements, on-site solar where practical, wind power, battery storage, waste-heat reuse, efficient chips, server virtualization, better software, and direct liquid cooling can reduce the impact. Operators can also use reclaimed water, select water-stable locations, extend hardware life, and publish clear data on electricity, water, and emissions.

Key takeaways

Data centers are essential digital infrastructure, but they are resource-intensive physical facilities.

A 1 MW IT data center operating at a PUE of 1.50 uses about 13.14 GWh of electricity each year. At Rs. 8 per kWh, that is roughly Rs. 10.51 crore in annual electricity cost.

As capacity rises from 1 MW to 100 MW, electricity use, water demand, operating costs, and emissions grow enormously. Efficient design helps, but clean electricity, responsible water planning, durable hardware, and transparent reporting matter just as much.

Disclaimer

The calculations in this article are illustrative estimates based on stated assumptions for PUE, electricity tariffs, grid emission factors, and water use. Actual data-center costs, power consumption, water needs, server capacity, and carbon emissions can vary by location, climate, hardware, cooling design, workload, electricity source, and local regulations. Readers should obtain site-specific technical, financial, and environmental assessments before making business, investment, or infrastructure decisions.

India’s startup funding scene is seeing interest in businesses that combine technology with real-world industries. Two recent stories highlight this shift. Materials-science startup Whizzo is reportedly in talks to raise $25 million from Bertelsmann India Investments, while biotech company Peptris Technologies has secured Rs. 86 crore from the Technology Development Board for AI-based drug development.

The two companies work in very different sectors, but both are trying to solve difficult problems through science, technology and specialized research.

Whizzo aims to modernize technical textile manufacturing

Whizzo was founded in 2024 by Shrestha Kukreja, a former Zetwerk executive.

The Bengaluru-based company develops technical textiles used in areas such as agriculture, infrastructure, packaging, protective clothing and industrial applications.

Unlike regular fabrics used mainly for comfort or fashion, technical textiles are designed for a specific purpose. They may need to resist heat, water, chemicals, fire or heavy pressure. For example, a geotextile can be used in road construction, while a protective textile may be needed for industrial workers.

Whizzo follows a contract development and manufacturing model. In simple terms, it helps customers move from an early product idea to research, testing, prototypes and large-scale manufacturing through one connected system.

According to earlier reports, Whizzo had raised $15 million in a Series A round led by Fundamentum. The money was intended to support materials research, intellectual property development and the expansion of its manufacturing network across Asia. The company had also planned to increase exports and work with more small and medium-sized manufacturers.

The proposed $25 million fundraise from Bertelsmann India Investments could help Whizzo expand its research facilities, build more proprietary textile products and increase production capacity. However, the round is still in discussion, so the final amount, valuation and use of funds may change as per Moneycontrol.

Peptris is taking AI drug discovery towards clinical development

Peptris Technologies was founded in 2019 by Venkatasubramanian Narayanan, Shridhar Narayanan, Anand Budni and Amit Mahajan.

The company uses artificial intelligence and machine learning to identify and improve potential drug molecules.

Its platform can study large chemical datasets, suggest new molecules and predict important properties such as safety, activity and suitability for further testing. This can reduce the number of compounds that need to be tested manually in laboratories.

The Rs. 86 crore support, approximately $8.98 million, has been approved under the government’s Research Development and Innovation Fund. It will support Peptris’s three-year VIGOR programme, which has a total project cost of Rs. 172 crore.

The programme will work on four AI-discovered drug candidates linked to conditions including neuromuscular disorders, cancer, alopecia, neuropathy and Duchenne muscular dystrophy. The plan is to move these candidates from an early validated stage towards advanced preclinical research and preparation for clinical trials as reported in PIB.

This is not a traditional equity investment. The support is structured as optionally convertible debt, which means the funding is provided as debt but may be converted into company shares under agreed conditions.

Disclaimer

The Whizzo fundraise is reportedly under discussion and should not be treated as a completed funding round until officially confirmed. Funding structures, amounts and deployment plans may change. This article is for information only and is not investment advice.

AI is creating a strong demand for computing power, and Indian startup Nava wants to build the infrastructure behind that growth. The company is reportedly in talks to raise around $200 million from existing investor Greenoaks.

The funding has not been finalized. According to Moneycontrol, discussions are still under way and the final size or terms could change.

If completed, the round could value Nava at nearly $750 million and become one of the largest recent investments in India’s AI infrastructure sector.

What is Nava’s core business

Nava, earlier known as Kluisz, is building private cloud infrastructure for companies that want to run artificial intelligence workloads securely and at scale.

In simple terms, Nava aims to provide the computing power, software and data centre systems needed to train and operate AI models. Instead of buying and managing expensive hardware on their own, businesses can use Nava’s infrastructure when they need it.

The company’s platform includes GPU computing, AI-focused data centres, orchestration software and tools for running AI models. GPUs are specialized computer chips that can process large amounts of data quickly, making them important for generative AI, machine learning and other demanding applications.

Nava’s official website describes the platform as being designed specifically for the AI era rather than adapting older cloud systems for new workloads. Its services are aimed at enterprises building AI products, running internal models and handling large-scale data operations. Nava itself says its focus is on secure, reliable and AI-ready cloud infrastructure.

Founders and company background

Nava was founded in 2025 by Abhinav Sinha, Vamshidhar Reddy and Abhijeet Singh.

Sinha previously worked as a global chief operating officer and chief product officer at OYO. Reddy is a former McKinsey partner and AMD executive, while Singh earlier worked as vice president of cloud infrastructure at Jio and also held a role at AT&T.

The team brings experience from cloud computing, consulting, large technology companies and business operations. That mix is important because an AI data centre startup needs more than software. It also needs knowledge of hardware, networks, energy use, reliability and enterprise sales.

How the $200 million could be used

Nava has not publicly disclosed a detailed spending plan for the proposed round. However, the company’s earlier funding announcements offer some clues.

Nava raised $22 million in a Series A round led by Greenoaks in April 2026. The company said that money would support GPU capacity, AI data centre development, specialized hiring and expansion across India and Southeast Asia. The Financial Express reported that Nava also planned to develop a full-stack AI cloud platform.

The larger proposed round could help Nava expand its GPU fleet, set up more computing capacity, hire data centre and AI engineering specialists, and build operations in new markets. It may also support research into better ways to manage AI workloads.

These are reasonable possibilities based on the company’s earlier plans, not confirmed details about the new round.

Competition in the AI cloud market

Nava operates in a growing market that includes NeevCloud, NxtGen, E2E Networks and Yotta Data Services. These companies also offer cloud, GPU or data centre services to businesses. Mint has identified these firms as competitors in India’s developing AI computing space.

Nava’s challenge will be to offer reliable performance at a price that companies can afford. It will also need enough hardware and data centre capacity to serve customers consistently.

Conclusion and key takeaways

Nava is trying to become an important infrastructure provider for the AI economy. Its work is less visible than a consumer app, but it supports the systems that allow AI applications to function.

The reported $200 million round is still being negotiated. If it closes, the money could help Nava expand its GPU capacity, strengthen its AI cloud platform and build a larger presence across India and Asia.

Key takeaways –

• Nava is reportedly in talks to raise around $200 million.
• The company was founded in 2025 by Abhinav Sinha, Vamshidhar Reddy and Abhijeet Singh.
• Its platform combines GPUs, cloud software and AI data centre infrastructure.
• The proposed funding has not yet been officially finalised.
• Likely uses include data centre expansion, GPU purchases, hiring and regional growth.
• Nava competes with other Indian cloud and AI infrastructure providers.

Zoho has introduced Zia Chat, a conversational AI tool designed to help employees work across business applications from one place. Instead of opening several apps to find information, users can ask questions in normal language and receive answers based on their authorized business data.

The announcement was made at a Zoho event in Chennai. Zoho’s co-founder and chief scientist Sridhar Vembu described the problem as an “interface crisis”, where employees spend too much time moving between software tools and searching for information. HT reported that Zia Chat is designed to bring these activities into one conversational interface.

What is Zia Chat

Zia Chat is more than a chatbot that answers general questions. It is designed to search business records, understand context, analyze information and complete selected tasks.

A sales manager could ask, “Which customers have not received a follow-up this week?” Zia Chat could search the relevant records and present the result. The user may then ask it to prepare a follow-up email or create tasks for the sales team.

Zoho says Zia Chat can help users search across applications, prepare reports and presentations, draft emails, create tasks and perform other business actions. It can also connect information between different Zoho products, depending on the employee’s existing access rights.

The tool is also expected to work with external applications through Model Context Protocol integrations. In simple terms, this can allow businesses to connect other software and bring information from those systems into the same chat experience.

Core capabilities of Zia Chat

The main strength of Zia Chat is its connection with business context. It is intended to understand information from tools such as CRM, email, analytics, customer support and workplace applications.

Its key capabilities include –

• Searching business data using everyday language
• Comparing sales, customer or financial information
• Creating reports and presentations
• Drafting emails and other workplace content
• Creating tasks and updating records
• Connecting information from Zoho and selected third-party apps
• Linking answers back to the records used
• Asking for confirmation before completing important actions

Zoho also says that Zia Chat will follow user permissions. This means an employee should only be able to view information or perform actions that their account already allows. Zoho has further stated that data accessed through Zia Chat will not be used to train its AI models. Zoho’s Zia chat official page also highlights its shared data model and company-owned technology stack.

What does ubiquitous intelligence mean

The phrase “ubiquitous intelligence” means AI that is available throughout daily work instead of being limited to one separate application.

For example, a customer support employee may ask Zia Chat to summarize a complaint, check the customer’s previous interactions and prepare a suitable response. A finance employee may ask for a comparison of monthly expenses. A manager may request a summary of pending tasks across departments.

The aim is to reduce repetitive searching and help employees make decisions faster. Zoho’s existing Ask Zia tools already support natural-language searches, data analysis, reports and recommendations across business systems. Zoho explains these capabilities on its official website.

Availability and competition

Zoho says Zia Chat is expected to enter early access within a month of its announcement. The initial version will be available in English under a bring-your-own-key model, which means businesses may need to provide their own AI model access. More languages are expected later.

Zia Chat will compete with other workplace AI assistants that are trying to connect email, documents, customer records and business software. Its main advantage could be the depth of its connection with the Zoho ecosystem. However, its usefulness will depend on how many third-party tools it supports, how accurate its answers are and how much control administrators receive.

Conclusion and key takeaways

Zia Chat shows Zoho’s wider plan to make AI a regular part of business software. Its purpose is not simply to generate text. The bigger goal is to help employees find information, understand data and complete work through a single conversation.

For a small business, this could mean fewer hours spent preparing reports or checking records. For a larger company, it could provide one common way to work across departments while still respecting access controls.

Key takeaways –

• Zia Chat is Zoho’s new conversational AI interface for business users.
• It can search information, analyse data, draft content and create tasks.
• It is designed to work across Zoho applications and selected external systems.
• User permissions and action confirmations are built into the experience.
• Zia Chat is part of Zoho’s wider vision of intelligence across everyday business software.

Image Credit- Zoho

BRAHMA AI has raised $150 million in a fresh funding round led by Multiples Alternate Asset Management. The deal has brought strong attention to the enterprise AI startup because it values the company at around $2 billion.

BRAHMA AI is trying to solve a very practical problem for big companies – how to create, manage, translate, secure, and reuse video and audiovisual content at scale. In simple words, it wants to become the AI operating system for enterprise content.

What BRAHMA AI does

BRAHMA AI works in the fast-growing area of audiovisual content. That includes video, digital humans, voice, multilingual content, and content intelligence.

For example, a global media company may need to turn one video into several language versions. A sports league may want digital fan experiences. A healthcare company may want safer training or communication tools using digital twins. BRAHMA AI is building tools that can support such use cases inside one enterprise-grade platform.

The company’s platform has two main parts – Brahma AI Core and Brahma AI Studio. Core focuses on content intelligence and management, while Studio is aimed at AI-powered creation. Together, they help businesses handle both the “organizing” and “creating” side of content.

Founders and company background

BRAHMA AI was established in 2024 and is founded and led by Prabhu Narasimhan, who serves as Founder and CEO. Jo Plaete is Co-founder and CTO, leading the company’s technology and R&D work.

The startup brings together capabilities developed across DNEG, Metaphysic, and Prime Focus Technologies. That background matters because these companies have experience in visual effects, media technology, and synthetic media. BRAHMA AI is now trying to convert that experience into a wider enterprise AI platform.

Funding details and purpose

The $150 million round was led by Multiples Alternate Asset Management. According to the company announcement, Multiples invested $100 million through preferred shares, and BRAHMA AI has also received another $100 million of investor interest.

The funding will mainly be used for three areas – research and development, global go-to-market expansion, and a stronger Silicon Valley presence. That means the company wants to build better AI products, hire and expand internationally, and get closer to major technology customers and partners in the US.

Business Standard reported that BRAHMA AI plans to expand its Bay Area presence and scale its global sales organization. This is important because enterprise AI is not only about good technology. Large clients also need support, trust, security, and long-term service.

Why this funding matters

The timing is interesting. Companies are no longer looking at AI only for chatbots or text generation. They want AI that can help with videos, ads, training content, healthcare communication, customer support, and localization.

BRAHMA AI is targeting sectors such as media and entertainment, sports, healthcare, and advertising. Its listed global customers and partners include Warner Bros., the NBA, Mayo Clinic, Google, Hakuhodo, and DNEG.

The startup is also working on interactive digital humans and a model-agnostic platform. Model-agnostic means the company does not want to depend on only one AI model. This can be useful for large companies that want flexibility, security, and better control over their data and workflows.

Competitors and market position

BRAHMA AI is entering a crowded but fast-moving market. It may compete with or overlap with companies such as Runway, Synthesia, Adobe, Luma AI, and other AI video or enterprise content platforms.

However, BRAHMA AI’s pitch looks slightly different. Instead of only offering creative AI tools, it is trying to become a full operating layer for enterprise audiovisual content. That means it wants to help companies create content, manage it, localize it, and use it safely across departments and markets.

Graph AI has raised $13.3 million in Series A funding, marking a fresh push for AI in the pharmaceutical safety space. The round was led by Insight Partners, with participation from existing investor Bessemer Venture Partners.

The startup is not building a regular chatbot or a general AI tool. Its focus is more specific. Graph AI wants to help pharma and biotech companies handle patient safety work faster, with better tracking and human review where it matters.

What Graph AI Does

Graph AI builds a platform called Graph Safety. It is made for pharmacovigilance, which simply means the process of tracking and reporting side effects, adverse events, and safety issues linked to medicines.

For example, if a patient reports a serious side effect after taking a medicine, a pharma company has to collect that information, check it, process it, and report it properly. This can take a lot of manual work, especially when reports come through emails, call centers, forms, doctors, hospitals, and other channels.

Graph Safety aims to bring these steps into one AI-supported system. The goal is not to remove safety experts, but to reduce repetitive tasks so they can spend more time on analysis and decisions.

Founders and Startup Background

Graph AI was founded in 2024 by Raghav Parvataraju, Vijay Ponukumati, Mohan Konyala, and Ashutosh Bordekar.

Raghav Parvataraju is the CEO of the company. The startup has positioned itself around patient safety operations for pharma and biotech firms, a market where accuracy, audit trails, and regulatory compliance are extremely important.

Funding Purpose

The new Graph AI Series A funding will mainly be used to expand the company across the United States and Europe. This is important because pharma safety work is global, and companies need systems that can support different markets and regulatory expectations.

Graph AI has already launched two modules. One helps capture and sort incoming adverse-event reports. Another works as an intelligent safety database for case processing. The company is also preparing a reporting module and is working on signal detection, which can help spot early safety patterns across cases.

In simple words, the funding gives Graph AI more room to build its product, grow its team, support enterprise customers, and enter bigger markets.

Why This Funding Matters

Patient safety is one of the less flashy but very important areas of healthcare. When medicine use grows, safety reports also increase. Many teams still depend on older systems, spreadsheets, manual checks, and disconnected tools.

This is where AI can be useful, but only if it is used carefully. In healthcare and pharma, companies cannot simply trust a black-box answer. They need proof, review trails, and human oversight.

Graph AI’s pitch is built around that idea. It wants AI to handle the heavy operational work, while experts remain in control of judgment and compliance.

Competitors and Market overview

Graph AI is entering a space with strong existing players. Some well-known names in pharmacovigilance software include Oracle Argus Safety, ArisGlobal LifeSphere Safety, and Veeva Vault Safety. Veeva has also announced Falcon Safety, an AI-focused safety operations product planned for early adopter availability in November 2026.

ArisGlobal and Veeva already serve large life sciences customers, so Graph AI will need to prove that its platform is faster, easier to adopt, and reliable in regulated environments. Its advantage may come from being AI-native from the start, instead of adding AI features on top of older workflows.

Conclusion and Key Takeaways

Graph AI’s $13.3 million Series A round shows that investors are looking closely at practical AI tools for healthcare operations, not just consumer-facing AI products.

Key takeaways –

• Graph AI raised $13.3 million in Series A funding led by Insight Partners.
• Bessemer Venture Partners also participated in the round.
• The startup was founded in 2024 and focuses on AI-powered patient safety software.
• The money will support expansion across the US and Europe.
• Graph AI competes with players like Oracle Argus, ArisGlobal LifeSphere Safety, and Veeva Safety.

Facts Input- PRNewswire

Meta has released Muse Glimmer, a new open-weight AI model designed to run on regular laptops and PCs. That one line is enough to make the launch interesting, because most powerful AI tools today still depend on cloud servers, paid APIs, or large data centres.

Muse Glimmer is Meta’s latest move in the open AI race. The company wants developers, students, researchers, and smaller businesses to have more control over how they use AI models. Instead of keeping everything locked behind a private system, Meta is again pushing the idea that more people should be able to download, study, and modify powerful AI tools.

With Muse Glimmer, we need to understand where AI is moving next. If useful AI can run directly on personal devices, laptops may become smarter work machines, not just screens connected to cloud apps.

What is Meta Muse Glimmer

Meta Muse Glimmer is a compact AI model from Meta’s Muse family. It is built to handle tasks such as reasoning, coding, planning, and everyday computer-based work.

In simple words, it is an AI model that can help with thinking and doing. It may help write code, organize files, draft emails, check information, plan workflows, or support small software tasks.

The most important part is that Muse Glimmer is described as an open-weight model that can run on a standard laptop or PC. That makes it different from many large AI systems that need expensive server hardware.

Open-weight does not always mean fully open-source. This is an important difference. In an open-weight model, the model’s trained parameters are shared so developers can download and use them. But the full training data, training method, or internal development process may not be completely public.

So, Muse Glimmer gives more freedom than a closed chatbot, but it may not mean everything behind the model is open.

Why running AI on laptops matters

Most people use AI through cloud-based tools. You type a prompt, the request goes to a server, and the answer comes back. This works well, but it has some limits.

First, it needs internet. Second, it can cost money at scale. Third, users may not want to send sensitive files, personal notes, or business documents to a cloud service.

A laptop-friendly AI model can reduce some of these problems. If the model runs locally, it can work faster for some tasks, give more privacy, and allow developers to build tools without paying for every single request.

For example, a small business owner could use a local AI assistant to sort customer queries, prepare invoices, or draft product descriptions. A student could use it to understand code without depending fully on an online chatbot. A developer could build a private helper for internal documents.

This is why Muse Glimmer is not just another model launch. It points toward a future where AI becomes part of the device itself.

How Muse Glimmer is connected to Muse Spark

According to reports, Muse Glimmer was developed using a method called distillation from Meta’s more powerful Muse Spark model.

Distillation is easier to understand with a classroom example. Imagine a senior teacher has deep knowledge, but it is not practical for every small school to hire that teacher. So the teacher’s knowledge is simplified into a useful guide that junior teachers and students can use more easily.

In AI, distillation means a smaller model learns from a larger model. The aim is to keep much of the usefulness while making the smaller model cheaper and easier to run.

Meta had earlier introduced Muse Spark as part of its Meta Superintelligence Labs work. Muse Spark 1.1 is designed for coding, tool use, computer tasks, and multimodal work, which means it can work with more than just text. Meta has also made Muse Spark available through the Meta Model API for developers in public preview.

Muse Glimmer looks like the lighter, more accessible member of this family.

Why Meta is pushing open-weight AI again

Meta has a long history with open AI models through its Llama series. The company gained strong developer attention by releasing models that people could download and build with.

With Muse Glimmer, Meta seems to be returning strongly to that approach. Mark Zuckerberg has argued that AI should not be controlled only by a few companies or governments. His recent essay, “The Future Is for Everyone,” presents Meta’s view that personal AI tools should be widely available.

This is also a business strategy. If many developers build apps using Meta’s models, Meta can become an important part of the AI ecosystem. It may not need to win every paid chatbot battle directly. It can still benefit if its models become widely used across apps, devices, and developer tools.

Who can use Muse Glimmer

Muse Glimmer may be useful for people who want more control over AI.

Developers can test it for local coding assistants, offline tools, automation, and private apps. Researchers can study how smaller models behave. Startups can build prototypes without depending completely on expensive cloud models. Power users can experiment with personal AI workflows on their own machines.

A practical example would be a freelance developer building a local assistant that reads project files and suggests fixes. Another example could be a small accounting firm using a private AI tool to summarize documents without uploading everything to a third-party server.

Of course, performance will depend on the laptop, memory, setup, and the model’s actual size. A normal office laptop may handle smaller tasks, while heavier use may need stronger hardware.

Competitors and market landscape

Meta is not alone in this race. OpenAI, Anthropic, Google, Mistral, DeepSeek, Alibaba, and several other companies are fighting for developer attention.

OpenAI and Anthropic mainly offer powerful closed models through apps and APIs. Google has Gemini models and strong cloud integration. Mistral has become known for open-weight models in Europe. DeepSeek and Alibaba’s Qwen models have pushed low-cost and open-weight AI from China into global developer discussions.

Meta’s advantage is its large user base, strong AI research teams, and history of developer adoption through Llama. Its challenge is trust. Some developers like open-weight models, but critics worry about safety, misuse, copyright, and the control Meta may still have over the larger ecosystem.

What this means for normal users

For everyday users, Muse Glimmer may not immediately feel like a new app they can download and use in one click. The first wave will likely matter more to developers and technical users.

But over time, laptop-ready AI models can change normal software. Your writing app may get a private assistant. Your coding tool may work better offline. Your personal finance app may analyze spending without sending data outside your device. Your laptop may handle more tasks even when the internet is weak.

That is the real promise of local AI – more useful tools, less waiting, and better control over private data.

Conclusion with key takeaways

Meta Muse Glimmer is important because it brings the AI conversation closer to personal devices. Instead of keeping advanced AI only in the cloud, Meta is showing that smaller, capable models can run on laptops and PCs.

It is still early, and users should wait for more technical details, real-world testing, and independent reviews. But the direction is clear. AI is moving from big servers toward personal machines.

Key takeaways –

• Meta Muse Glimmer is an open-weight AI model built for laptops and PCs.
• It is designed for reasoning, coding, planning, and computer-based tasks.
• Open-weight means developers can access model weights, but it is not always the same as fully open-source.
• Muse Glimmer is reportedly distilled from Meta’s stronger Muse Spark model.
• The model could help developers build more private, local AI tools.
• Meta is competing with OpenAI, Anthropic, Google, Mistral, DeepSeek, Alibaba, and others in the AI model race.

Facts Input- BI, AP News, The Guardian

Artificial intelligence is now part of daily life, even for people who do not work in technology. It recommends videos, unlocks phones, checks spam emails, helps banks detect fraud, and powers chatbots that answer questions in seconds.

But the words around AI can be confusing. One person says AI, another says Gen AI, and someone else talks about AGI as if it is already here. These terms are related, but they do not mean the same thing.

The simple way to understand it is this. AI is the big umbrella. Gen AI is one important type of AI that creates new content. AGI is a future idea where machines could think and learn across many tasks like humans. Let’s break this down in simple words.

What Is AI (Artificial Intelligence)

AI, or artificial intelligence, means computer systems that can do tasks that normally need human thinking. These tasks may include learning, predicting, recognizing images, understanding speech, making recommendations, or helping with decisions.

A simple example is Google Maps. When it suggests the fastest route, it studies traffic, road conditions and past travel patterns. That is AI at work.

Another example is online shopping. When Amazon or Flipkart recommends a product based on your search and buying history, it is using AI to guess what you may like.

AI does not always create something new. Many AI systems are made to classify, predict or decide. For example, a bank may use AI to detect a suspicious transaction. A hospital may use AI to help doctors study medical scans. A car may use AI-based systems to detect lanes or warn the driver.

Most AI used today is narrow AI. That means it is built for a specific job. It may be very good at that job, but it cannot suddenly behave like a human across every subject.

What Is Gen AI (Generative Artificial Intelligence)

Gen AI, or generative AI, is a type of artificial intelligence that can create new content. This content can be text, images, videos, music, computer code, presentations or summaries.

ChatGPT, Gemini, Claude, Midjourney and similar tools are common examples of Gen AI. You give them a prompt, and they generate an answer or output.

For example, if you ask a Gen AI tool to write an email, it can create a draft. If you ask it to make a product description, it can produce one. If you ask an image model to create a picture of a futuristic electric scooter, it can generate an image based on patterns it learned from training data.

The main difference is that Gen AI produces something new. Traditional AI may tell you whether an email is spam. Gen AI can write the email itself.

But Gen AI is not magic. It does not “know” things in the same way a person does. It works by learning patterns from huge amounts of data and then creating outputs that match the request. This is why it can sometimes make mistakes, sound confident when wrong, or invent details. Human checking is still important.

What Is AGI (Artificial General Intelligence)

AGI stands for artificial general intelligence. It refers to a possible future type of AI that could understand, learn and solve many different kinds of problems at a human-like level.

Today’s AI tools are powerful, but they are still limited. A chatbot can write a report, explain a topic or help with code. But it does not truly live in the world like a human. It does not learn every new task naturally from experience in the same broad way people do.

AGI would be different. In theory, it could move from one type of task to another without needing separate training for each one. It could reason across subjects, adapt to new situations, and apply learning from one area to another.

A simple example makes this easier. A human can learn cooking, driving, writing, planning a trip and solving a workplace problem. These tasks are very different, but people can connect experience across them. AGI would aim for that kind of flexible intelligence.

At present, AGI is still a research goal, not a confirmed reality. Experts also do not fully agree on how exactly AGI should be defined or measured.

AI vs Gen AI vs AGI – The Simple Difference

AI is the broad field. It includes many systems that can learn, predict, recommend or decide.

Gen AI is a smaller part of AI. It focuses on creating new content from prompts or inputs.

AGI is a future-level concept. It describes AI that could match or exceed human ability across a wide range of thinking tasks.

Think of it like transport. AI is the full transport system. Gen AI is one type of vehicle inside that system. AGI is like imagining a vehicle that can drive, fly, sail, repair itself, learn new routes and handle almost any travel problem without being redesigned.

That is why these terms should not be used as if they are the same.

Everyday Examples

AI example – Your phone camera detects faces and adjusts focus.

Gen AI example – An app creates a birthday invitation design from your text prompt.

AGI example – A future system could plan the birthday party, compare budgets, design the invite, book the venue, handle guest questions, adjust the plan after rain, and learn from the whole experience like a human assistant. This kind of system does not fully exist today.

Where AI Is Used Today

AI is already used in finance, healthcare, education, retail, transport, manufacturing and mobile apps.

In finance, AI can help detect fraud, study credit risk and improve customer support. In automobiles, it supports driver assistance features, parking systems and route planning. In mobile phones, AI improves photos, battery use, voice typing and app suggestions.

These systems are useful because they process large amounts of information faster than humans. But they work best when the task is clear and the data is reliable.

Where Gen AI Is Used Today

Gen AI is popular because it helps with content and creativity. Writers use it for outlines. Developers use it to explain or draft code. Marketers use it for campaign ideas. Students use it to understand topics. Companies use it for chatbots, reports, summaries and design support.

In business, Gen AI can save time on repetitive writing or research tasks. For example, a customer support team can use it to draft replies faster. A sales team can use it to summarize client calls. A finance team can use it to prepare first drafts of reports.

Still, Gen AI should be used carefully. It can produce wrong facts, biased answers or copied-sounding text if used lazily. The best results come when humans review, edit and verify the output.

Why AGI Gets So Much Attention

AGI gets attention because it could change how work, science and technology function. If a machine could learn broadly like a person, it might help solve complex problems in medicine, climate research, engineering and education.

But AGI also brings serious questions. Who controls it? How safe would it be? How would it affect jobs? Could it make decisions that humans do not understand?

This is why many researchers and organizations talk about AI safety, testing and responsible development. The more powerful AI becomes, the more important it is to manage risks properly.

Common Confusion Around These Terms

Many people call every smart tool “AI,” which is understandable but not always accurate. A calculator is not AI. A simple rule-based tool may not be AI either. AI usually means the system can use data or patterns to make predictions, decisions or outputs.

Another common confusion is thinking Gen AI is the same as AGI. It is not. Gen AI can sound smart because it produces fluent answers, but that does not mean it has human-level general intelligence.

A chatbot may explain physics in one moment and write a poem in the next, but it can still fail at basic reasoning or make up a source. That gap is important.

Conclusion With Key Takeaways

AI, Gen AI and AGI are connected, but they are not interchangeable. AI is the wider technology field. Gen AI is the creative branch that generates new content. AGI is a possible future stage where machines could show broad, human-like intelligence.

For most people and businesses, the practical focus today should be AI and Gen AI. These tools are already useful in mobile apps, finance, automobiles, customer service, education and content creation. AGI is worth understanding, but it is still a developing idea rather than an everyday product.

Key takeaways –

• AI is the broad term for machines that can learn, predict, recommend or make decisions.
• Gen AI is a type of AI that creates new content such as text, images, video or code.
• AGI means human-like general intelligence across many tasks, but it has not been clearly achieved yet.
• Most AI tools today are narrow AI, meaning they are built for specific tasks.
• Gen AI is powerful, but its output should still be checked by humans.

Facts Input- 1, 2, 3

Discovered Materials has raised $9 million in seed funding at a time when the chip industry is looking for answers beyond just smaller transistors and faster processors. The startup is working on a hard but important problem – finding new materials that can make semiconductor chips more efficient, cooler and easier to scale.

The seed round was led by Lightspeed India Partners. Y Combinator, Peak XV Partners and angel investors including Paul Graham, Gokul Rajaram and Thariq Shihipar also participated. As per reports, the funding is about Rs. 85 crore.

At the heart of the company’s plan is a simple idea. If AI is creating a huge demand for more computing power, then AI can also help find the materials needed to support that demand.

What Discovered Materials Is Building

Discovered Materials is building AI agents for semiconductor material discovery. In simple terms, these AI agents help researchers search for new materials that may be useful in chips, data centres and semiconductor factories.

Finding a useful new material is not like testing a new app feature. It can take years of lab work, failed experiments and careful validation. A material may look promising on paper but fail when it is tested in real conditions. It may also be too expensive or difficult to manufacture at scale.

Discovered Materials wants to shorten this long journey. The company says its AI agents work across simulation, synthesis and experimental testing. That means the system does not only suggest ideas. It is meant to help move promising material candidates closer to real lab validation.

Why The Funding Matters

The $9 million seed round gives Discovered Materials more room to build its team, expand lab work and improve its AI research agents.

This is important because the startup is working in a space where software alone is not enough. It needs strong AI models, but it also needs scientists, lab equipment and industry partnerships. A new chip material has to prove itself in the physical world.

The funding will also help the company work faster on thermal materials. These are materials that help manage heat. Heat is one of the biggest problems in modern AI chips. As chips become more powerful, they also produce more heat, and that heat needs to be moved away efficiently.

For a simple example, think about how your phone heats up during gaming or video recording. Now imagine that problem inside a data centre running thousands of AI chips all day. If heat is not managed well, performance drops, energy use rises and cooling costs go up.

Founders And Background

Discovered Materials was founded in 2026 by Akash Ramdas and Advaith Sridhar. The company is listed as a Y Combinator Spring 2026 startup and is based in San Francisco.

Both founders have links to IIT Madras, where they met more than a decade ago. Their backgrounds also explain why the company is trying to mix material science with AI.

Akash Ramdas has a PhD in Materials Science from Stanford University and has spent years researching new materials for semiconductor chips. His work includes nanoscale interconnect materials, which are important for how tiny chip components connect and transfer signals.

Advaith Sridhar studied AI at Carnegie Mellon University and has worked on AI models and agents at companies including Persona AI and Luma Labs. His experience fits the company’s AI-first approach to scientific research.

Together, they are trying to solve a problem that sits between two worlds – deep science and modern AI.

The Main Purpose Of The Funding

The aim of the funding is not just to grow quickly like a normal software startup. Discovered Materials needs capital to build a research engine that can move from idea to lab-tested material faster than traditional methods.

There are three clear purposes behind this seed round.

1. First, the company wants to scale its AI agents. These agents need to understand scientific problems, generate useful material ideas and help researchers make better decisions.
2. Second, the company wants to expand lab work. AI predictions are useful only when they survive real testing. A material that works in a simulation still has to be made, measured and compared with existing options.
3. Third, Discovered Materials wants to support the chip industry’s need for better thermal performance. As AI chips become hotter and more power-hungry, better heat management can become a major advantage.

Why Semiconductor Materials Are So Important

Most people hear about chips through names like Nvidia, AMD, Intel, TSMC or Samsung. But behind every chip is a long list of materials that affect speed, heat, power use and reliability.

Materials are used in packaging, wiring, insulation, cooling layers and many other parts of the chip system. If these materials improve, the chip can become more efficient without changing everything else.

This matters because chip progress has become harder. For years, the industry benefited from making transistors smaller. That path is now more difficult and costly. So companies are looking at other ways to improve performance. Better materials are one of those ways.

Discovered Materials is trying to sit right at this pressure point.

Material Discovery Bench

Along with the funding news, Discovered Materials has introduced Material Discovery Bench. The company describes it as an open-source benchmark for testing how well AI models can handle real semiconductor material discovery problems.

This is useful because AI tools often sound impressive, but scientific research needs proof. A benchmark gives researchers a way to compare models and check whether they are actually useful for serious material problems.

According to the company, the benchmark was built with inputs from experts connected to institutions and companies such as IBM, IMEC, Stanford and Cambridge. It is expected to grow further with simulation and experimental validation.

Competitors And Similar Companies

Discovered Materials is part of a growing group of startups using AI to speed up materials research.

• Materials Zone is one such company. It works on materials data and AI tools that help research teams manage and learn from experimental data.
• Materials Nexus, based in the UK, uses AI and physics-based methods to discover cleaner and more efficient materials. It has worked on areas linked to climate technology, batteries and semiconductors.
• Orbital Industries is another company in the broader AI materials space, with work linked to data centre infrastructure and advanced materials.

These companies do not all solve the same problem. Some focus on data systems, some on clean energy materials, and some on industrial hardware. But they show that AI-led material discovery is becoming a serious startup category.

Conclusion With Key Takeaways

Discovered Materials’ $9 million seed round is not just another AI funding story. It points to a bigger shift in the semiconductor industry. Chips need better materials, especially as AI workloads create more heat and demand more power.

The company’s bet is that AI agents can help scientists find and test new materials much faster than older research methods. If that works, it could help chipmakers improve performance, reduce heat problems and move useful discoveries from lab to fab more quickly.

Key takeaways –

• Discovered Materials has raised $9 million, about Rs. 85 crore, in seed funding.
• The round was led by Lightspeed India Partners, with participation from Y Combinator, Peak XV Partners and angel investors.
• The startup was founded in 2026 by Akash Ramdas and Advaith Sridhar.
• Its main focus is using AI agents to discover new semiconductor materials.
• The funding will support team growth, lab expansion and stronger AI research agents.

Facts Input- BS

Artificial Intelligence is no longer a future topic. It is already part of banking apps, shopping recommendations, customer support, fraud detection, healthcare tools, coding assistants, cameras, chatbots and business software.

For students and working professionals, this creates a strong career opportunity. But it also creates confusion. Many people want to become an AI engineer, but they do not know where to start, what to learn, or which job role to target first.

The good news is that AI has many career paths. You do not need to become a research scientist on day one. You can start with programming, data, machine learning, software engineering or cloud, and slowly move toward advanced AI roles.

What does an AI engineer do

An AI engineer builds systems that can learn from data, make predictions, understand language, recognize images, automate tasks or support decision-making.

For example, an AI engineer may build a model that detects fake transactions for a bank. Another may create a chatbot for customer support. Someone else may build a recommendation engine for an ecommerce app or an AI tool that helps doctors analyze medical reports.

In simple words, an AI engineer turns data and algorithms into useful products.

This role is different from only using AI tools. An AI engineer understands how models work, how to train them, how to test them and how to deploy them in real applications.

Why AI careers are growing

AI skills are becoming important across industries.

The World Economic Forum’s Future of Jobs Report 2025 lists AI and big data among the fastest-growing skills. NIIT’s India Skills Gap Report 2026 also highlights AI, digital, data and cybersecurity skills as key future capabilities in India.

Indian companies are hiring AI talent for fintech, healthcare, retail, IT services, manufacturing, telecom, education, cybersecurity and SaaS products.

AI may automate some tasks, but it is also creating new roles. Companies need people who can build, manage, secure and improve AI systems.

Main career options in Artificial Intelligence

AI engineer –

This is one of the most direct roles. An AI engineer builds AI models and integrates them into real products.

They work with machine learning, deep learning, Python, APIs, cloud platforms and data pipelines. Their job is to make AI useful for business problems.

Best for – People who like coding, maths and product building.

Machine learning engineer –

A machine learning engineer focuses on models that learn from data.

They may build systems for fraud detection, customer prediction, pricing, image recognition or recommendation. They also work on model testing, accuracy and deployment.

Best for – People who enjoy programming and data-driven problem-solving.

Data scientist –

A data scientist studies data to find patterns and build predictive models.

For example, a data scientist at a telecom company may predict which customers are likely to leave. A retail data scientist may study which products sell better in which city.

Best for – People who like statistics, business questions and storytelling with data.

AI researcher –

AI researchers work on new algorithms, model architectures and advanced AI methods.

This role usually needs strong mathematics, deep learning knowledge and often a master’s or PhD. It is common in research labs, top tech companies and advanced AI startups.

Best for – People who enjoy theory, experiments and long-term research.

Generative AI engineer –

This is a newer role. Generative AI engineers build applications using large language models, image models, voice models and multimodal AI.

They work on chatbots, copilots, AI agents, summarization tools, document assistants and content automation systems.

Best for – People who like building practical AI tools using models like GPT, Claude, Gemini, Llama or other open-source models.

MLOps engineer –

MLOps means machine learning operations.

An MLOps engineer makes sure AI models run properly after deployment. They handle model monitoring, versioning, cloud infrastructure, automation and reliability.

For example, if a fraud detection model becomes less accurate over time, MLOps systems help detect the problem.

Best for – People who like cloud, DevOps and automation.

Computer vision engineer –

Computer vision engineers build AI systems that understand images and videos.

Their work is used in medical scans, self-driving systems, factory inspection, face recognition, drones, agriculture and retail analytics.

Best for – People who like image processing, cameras and visual AI.

NLP engineer –

NLP means natural language processing. NLP engineers build systems that understand text and speech.

They work on chatbots, translation, voice assistants, sentiment analysis, search tools and document understanding.

Best for – People interested in languages, text, speech and communication.

AI product manager –

An AI product manager does not always build models directly. They decide what AI product should be built, why users need it and how it should work.

They connect engineering, design, business and customers.

Best for – People who understand technology but also enjoy strategy and communication.

Skills needed to become an AI engineer

1. The first skill is programming. Python is the best starting point because it is widely used in AI.
2. The second skill is mathematics. You do not need to become a maths professor, but you should understand basic linear algebra, probability, statistics and calculus.
3. The third skill is data handling. Learn SQL, Pandas, NumPy and data cleaning.
4. The fourth skill is machine learning. Learn supervised learning, unsupervised learning, classification, regression, model evaluation and feature engineering.
5. The fifth skill is deep learning. Learn neural networks, CNNs, RNNs, transformers and model training.
6. The sixth skill is deployment. An AI model is useful only when people can use it. Learn APIs, Docker, cloud basics and MLOps tools.
7. The seventh skill is communication. AI engineers must explain model results, limitations and risks to non-technical teams.

Beginner roadmap for AI engineer

1. Start with Python. Build small projects like a calculator, file organizer or data cleaner.
2. Then learn data analysis using Excel, SQL, Pandas and visualization tools.
3. After that, study machine learning basics. Build projects like house price prediction, spam detection or customer churn prediction.
4. Next, learn deep learning with PyTorch or TensorFlow. Try image classification or text classification projects.
5. Then move to generative AI. Build a document summarizer, chatbot, resume analyzer or AI search tool.
6. Finally, learn deployment. Put your project on GitHub and deploy it using FastAPI, Streamlit, Docker or a cloud platform.

Practical projects to build

1. A resume screening tool for recruiters.
2. A chatbot for college FAQs.
3. A fraud detection model for payment data.
4. A crop disease detection app using plant images.
5. A customer review sentiment analyzer.
6. A PDF summarizer for students.
7. A recommendation system for movies or products.
8. A voice-based assistant for local language users.

These projects show employers that you can solve real problems, not only complete courses.

Career path for AI engineer

The path usually starts with junior roles.

You may begin as a data analyst, Python developer, machine learning intern, AI intern or junior data scientist.

After one to three years, you can move into machine learning engineer, AI engineer, NLP engineer or computer vision engineer roles.

After five or more years, you can become senior AI engineer, MLOps lead, AI architect, data science manager or AI product leader.

With deeper research skills, you can move toward applied scientist or AI researcher roles.

Who can enter AI

Students from computer science, IT, electronics, mathematics, statistics and engineering backgrounds have a natural advantage.

But people from other fields can also enter AI if they build the right skills. Commerce students can move into business analytics. Healthcare professionals can work in health AI. Finance professionals can move into risk modelling. Language students can work in NLP and AI evaluation.

AI rewards people who combine technical skills with domain knowledge.

Common mistakes to avoid

• You should not start with only ChatGPT prompts and call yourself an AI engineer. Prompting is useful, but it is not enough.
• You should not ignore maths completely. You need at least basic understanding.
• You should not collect too many certificates without projects.
• You should not copy projects blindly from YouTube. Build something small but original.
• You should not chase every new AI tool. Focus on fundamentals first.

Conclusion with key takeaways

Artificial Intelligence offers many career options, from AI engineer and data scientist to MLOps engineer, NLP engineer, computer vision engineer and AI product manager.

For beginners, the best path is simple. Learn Python, understand data, study machine learning, build projects and then move into deep learning and generative AI. You do not need to master everything at once. You need steady learning and proof of work.

Key takeaways –

• AI engineering is one of the fastest-growing tech career paths.
• Python, data analysis, machine learning and deployment are core skills.
• Generative AI has created new roles in chatbots, AI agents and document automation.
• Projects matter more than certificates alone.
• The best AI careers combine technical skill with real-world problem understanding.

Facts Input- IE, ToI

Uttar Pradesh is preparing a major technology push with a planned robotics and AI manufacturing hub in Noida. The project is being presented as India’s largest integrated robotics and advanced manufacturing cluster, and the idea is much bigger than building one more industrial park.

The hub, called PRAGATI, is expected to bring together robotics, artificial intelligence, GPU clusters, testing labs, manufacturing facilities, startups and research teams in one place. For a state that already has a strong electronics and mobile manufacturing base around Noida and Greater Noida, this is a clear attempt to move up the value chain.

In simple words, UP does not want to only assemble devices. It wants to help companies design, test and manufacture the intelligent machines that may run future factories, hospitals, warehouses, airports, farms and defence systems.

What exactly is planned in Noida?

Chief Minister Yogi Adityanath has approved the plan to develop PRAGATI in Noida. PRAGATI stands for Park for Robotics, Artificial Intelligence, GPU Clusters and Advanced Technical Innovation.

The proposed hub will be spread across about 75 acres. According to reports, it will include robotics testing and certification facilities, rapid prototyping units, manufacturing infrastructure, AI compute capacity, GPU clusters, motion capture labs, physical AI data centres, startup incubation facilities and co-working spaces.

That sounds heavy, so let’s break it down.

A robotics company cannot grow only with a laptop and a coding team. It needs hardware, sensors, motors, cameras, test spaces, simulation tools and manufacturing support. A robot made for a warehouse, for example, must be tested for movement, safety, battery life, load handling and navigation. The Noida hub is being planned as a place where such companies can build and test products faster.

Why does UP need this hub?

India has a large market for automation, but much of the high-end robotics and AI hardware still depends on imports or foreign technology. This creates a gap. Indian companies may use automation, but fewer Indian firms build complete advanced robotic systems at scale.

UP wants to use Noida’s existing electronics ecosystem to close part of this gap. The region already has mobile phone manufacturing, component suppliers, expressway access, industrial land and proximity to Delhi-NCR’s talent pool. Adding robotics and AI manufacturing to this base is a natural next step.

The need is also linked to jobs. Reports say the project could create more than one lakh direct and indirect jobs and add over Rs. 2,000 crore in gross value over the next five years. These jobs may not be limited to engineers. A robotics ecosystem also needs technicians, machine operators, assembly workers, quality testers, logistics staff, trainers and support teams.

The aim and purpose of the initiative

The main aim is to make Uttar Pradesh a serious player in deep technology. Deep tech simply means technology built on serious engineering and research, not just a simple app or website.

The purpose is also industrial. Robots and AI systems are now becoming part of manufacturing, healthcare, logistics, agriculture, defence, surveillance, education and smart city projects. If UP can attract manufacturers and research centres early, it can become a supplier to these fast-growing sectors.

There is another important purpose – reducing the gap between research and real products. India has strong engineering talent, but many ideas struggle to move from lab to factory. A cluster with prototyping, testing, certification and manufacturing support can help startups and companies make that jump.

For example, a startup building a delivery robot may need a test track, sensors, metal fabrication, software support and safety certification. If all of this is scattered across different cities, the process becomes slow and expensive. A common hub can make the journey easier.

Why Noida makes sense

Noida is already one of North India’s strongest technology and electronics regions. Greater Noida and the Yamuna Expressway belt have attracted mobile manufacturers, electronics companies and data centre investments.

Invest UP says Uttar Pradesh accounts for a large share of India’s mobile phone manufacturing and has more than 230 operational electronics system design and manufacturing companies. Greater Noida is also home to major phone makers such as Samsung, OPPO and Vivo.

This matters because robotics is not only about software. It needs electronics, chips, sensors, cameras, boards, motors, batteries, casings and precision parts. A region that already understands electronics manufacturing has a better chance of building a robotics supply chain.

The upcoming Jewar airport may also help in the long run by improving cargo and global connectivity, especially if companies want to export components or finished products.

What kind of companies may come?

The hub is expected to attract global manufacturers, component suppliers, research and development centres and specialized technology service providers.

This could include companies working on industrial robots, warehouse automation, drones, autonomous machines, AI cameras, robotic arms, smart factory systems, defence robotics and healthcare robots.

Startups may also benefit. A young robotics startup usually struggles with cost. Buying testing equipment or setting up a small manufacturing line is expensive. If the hub offers shared facilities, incubation and technical support, it can reduce the early burden.

Bigger companies may use the hub for manufacturing, testing, pilot projects or partnerships with local suppliers.

How this links with UP’s wider AI plan

The Noida robotics hub is not a standalone idea. UP has also been working on AI City in Lucknow, the UP AI Mission, electronics manufacturing clusters, semiconductor-related opportunities and centres of excellence.

Invest UP’s AI City plan for Lucknow includes incubation, AI-focused workspaces, high-performance compute capacity, testing labs and simulation labs. The state has also talked about expanding the technology ecosystem beyond Noida and Ghaziabad into tier-II and tier-III cities.

This is important. If all new technology investment stays only in one belt, the impact remains limited. A wider network can help universities, startups and local businesses across the state connect with the deep-tech economy.

Competitors and other states to watch

UP is not the only state trying to become a technology hub. Karnataka is pushing hard on AI, with plans around AI education, innovation and data centres. Andhra Pradesh has been in the news for large AI and data infrastructure investments around Visakhapatnam. Telangana, Tamil Nadu, Gujarat and Maharashtra also compete for electronics, semiconductor, data centre and advanced manufacturing investments.

Noida’s advantage is its existing electronics base and NCR location. Its challenge will be execution. Announcing a hub is easier than building one that companies actually use every day.

The state will need quick approvals, reliable power, skilled workers, good testing facilities, strong industry partnerships and clear rules. Robotics companies will not come only for land. They will come if the ecosystem saves them time and money.

Challenges ahead

1. The first challenge is talent. Robotics needs mechanical engineering, electronics, embedded systems, AI, computer vision and manufacturing skills. UP will need strong training and certification programmes, as the government has already indicated.
2. The second challenge is supply chain depth. A robot is made of many parts. If key components still need to be imported, local manufacturing may remain limited.
3. The third challenge is real demand. Companies will invest only if they see buyers in factories, hospitals, logistics, defence, agriculture and public services.
4. The fourth challenge is patience. Robotics is not quick commerce. Products take time to build, test, certify and commercialize. The state must support companies beyond the launch phase.

Conclusion with key takeaways

The UP robotics and AI manufacturing hub in Noida could become an important project if it moves from announcement to strong execution. It brings together the right themes – AI, robotics, manufacturing, startups, GPU infrastructure and testing support.

For India, the bigger message is clear. The next phase of technology growth will not be only about apps. It will also be about machines, sensors, chips, automation and physical AI systems that work in the real world.

Key takeaways

• UP plans to develop PRAGATI, a 75-acre robotics and AI manufacturing hub in Noida.
• The project is being presented as India’s largest integrated robotics and advanced manufacturing cluster.
• It may include testing labs, GPU clusters, AI compute, prototyping units, data centres, manufacturing facilities and startup spaces.
• The hub is expected to create more than one lakh direct and indirect jobs and add over Rs. 2,000 crore in gross value over five years.
• Noida’s existing electronics and mobile manufacturing base gives the project a practical starting point.

Facts Input- Tele net, Startup UP, UPLC, Invest UP,

Vorflux, a new AI startup founded by Rippling co-founder Prasanna Sankar, has raised $15 million in seed funding to build what it calls an “autopilot” for software engineering.

The round includes backing from Y Combinator, Peak XV Partners, Powerset and Alliance, along with angel investors such as Parker Conrad, Immad Akhund and Balaji Srinivasan.

The startup is entering one of the hottest areas in AI right now – coding agents. But Vorflux is not pitching itself as another tool that only writes code snippets. Its bigger idea is to automate more of the software development cycle, from understanding a task to planning, building, testing and creating a merge-ready pull request.

What Vorflux does

Vorflux is building a cloud-based AI coding agent for software teams.

A developer or product team can give it a problem, such as fixing a bug, building a feature or improving an internal workflow. Vorflux then studies the codebase, prepares a plan, writes the code, runs it on a real machine, tests the output and creates a pull request that can be reviewed and merged.

In simple words, it wants to behave less like a junior assistant and more like an autonomous engineer that can take a task closer to completion.

The company’s own positioning is clear – “autopilot for software engineering.” That means it wants AI to handle the long, repetitive engineering work while humans focus more on judgment, product direction and final review.

Founder and background

Vorflux was founded by Prasanna Sankar, best known as a co-founder of Rippling, the HR and workforce management software company started in 2016.

Sankar remains associated with Rippling as its Chief Technology Officer. Vorflux is his new AI venture after his earlier Web3 startup 0xPPL shut down.

His background matters because Vorflux is not being built by someone outside software engineering. It is coming from a founder who has seen how large engineering teams work, where time gets wasted, and why shipping software is often slower than writing code itself.

Funding details

Vorflux has raised $15 million in seed funding.

The investors include Y Combinator, Peak XV Partners, Powerset and Alliance. Angel investors include Parker Conrad, who co-founded Rippling, Immad Akhund of Mercury, and Balaji Srinivasan, former CTO of Coinbase.

The money is expected to help Vorflux expand its engineering team, improve the product and build its go-to-market strategy for enterprise software teams.

Startup aim and purpose

Vorflux’s aim is to reduce the bottlenecks in software engineering.

Today, many AI tools can generate code. But coding is only one part of engineering. A real task also needs planning, environment setup, testing, code review, debugging, deployment checks and team-specific standards.

That is the gap Vorflux wants to solve.

For example, if a company asks an AI coding tool to “add billing export support,” the hard work does not end after the first code draft. Someone must check the database, understand existing billing logic, update tests, make sure the UI works, handle edge cases and open a clean PR.

Vorflux wants to take more of that full journey into one AI-driven workflow.

How it is different from normal coding assistants

Most coding assistants work inside a developer’s editor. They suggest code, complete lines or answer questions.

Vorflux is trying to work at a higher level. It runs in the cloud, connects to tools like repos, databases, CI/CD systems, error tracking, project management tools and communication apps, then works across the engineering workflow.

It also supports multiple AI models from different labs instead of depending on only one model. This matters because AI models are improving quickly, and the “best” model for coding can change from month to month.

Vorflux’s idea is that the model is only one part of the system. The real value comes from the workflow around the model – planning, testing, reviewing and remembering how a team works.

Why this matters now

Software teams are already using AI, but many are still stuck at the assistant stage.

A developer may ask AI to write a function, but still has to test it locally, fix broken parts, check standards and handle deployment. That saves some time, but not enough to change how teams ship software.

Vorflux is betting that AI has reached a point where it can handle longer engineering tasks, especially when it has the right cloud environment and review system.

If this works, the role of engineers may shift. Instead of writing every line themselves, they may spend more time defining the task, reviewing decisions, setting standards and checking whether the output is safe to ship.

Competitors

Vorflux enters a crowded and fast-moving market.

Its competitors and adjacent players include GitHub Copilot, Cursor, Replit, Cognition’s Devin, OpenAI Codex-style agents, Google’s AI coding tools, Anthropic’s Claude-based coding workflows and other agentic development platforms.

The difference will come down to reliability. Many tools can impress in demos. Fewer can consistently work on messy real codebases, run tests, follow company standards and produce merge-ready work.

That is the real battlefield.

Challenges ahead

Vorflux has a strong story, but it also has a hard problem to solve.

Enterprise codebases are complex. They have old systems, undocumented logic, security rules, internal tools and team habits that are not always written down. An AI agent must understand all this before making changes.

Trust is another challenge. Companies will not let an AI freely change production systems unless they believe the tool is safe, auditable and easy to control.

Cost may also matter. Running cloud machines and using frontier AI models can become expensive. Vorflux will need to show that its tool saves enough engineering time to justify the bill.

Conclusion – Key takeaways

Vorflux’s $15 million seed funding shows how quickly AI coding is moving from simple assistance to autonomous engineering workflows.

Founded by Rippling co-founder Prasanna Sankar, the startup wants to build an AI “autopilot” that can plan, code, test and ship merge-ready pull requests.

The purpose is not just to write code faster. It is to remove the slow parts of software delivery. If Vorflux can make that reliable for real enterprise teams, it could become an important name in the next wave of AI software development.

Facts Input- ET

Othor AI has raised Rs. 60 lakh in a bridge funding round ahead of its planned seed round. For an early-stage AI startup, this kind of bridge capital can be useful because it gives the company time to improve the product, test customer demand and prepare for a larger institutional raise.

Bridge rounds are common in startups. They usually sit between two bigger funding stages. In Othor AI’s case, the round appears to be a small but important step before the company goes out for seed funding.

The bigger question is not only how much money the startup has raised. The more interesting part is what Othor AI is trying to build, why AI startups need early capital, and how such funding can help a young company move from idea to market.

What Othor AI does

Unni Koroth and Nekender Shekhawat founded it in September 2024.

Othor AI is an artificial intelligence startup working in the fast-growing AI products and solutions space. While full product details are still limited in the public domain, the company’s direction is linked to building AI-led tools that can help users or businesses solve work faster.

In today’s market, AI startups usually focus on areas such as automation, analytics, content workflows, customer support, enterprise productivity, data processing or task-specific AI agents.

For Othor AI, the purpose of this bridge round is likely to support early product development and market validation before a larger seed fundraise.

Funding details

Othor AI has raised Rs. 60 lakh in a bridge round ahead of seed funding.

This amount may look small compared with large AI funding rounds, but for an early-stage startup it can be meaningful. It can cover product development, cloud costs, engineering work, pilots, customer testing and basic hiring.

AI startups often need more early money than normal software startups because they may use paid AI models, cloud infrastructure, datasets, testing tools and specialized technical talent.

Why bridge funding matters

A bridge round gives a startup breathing room.

Many young startups reach a stage where they have a working idea, early users or a prototype, but they are not yet ready for a large seed round. Investors may want more proof. Customers may want better features. The product may need more testing.

That is where bridge capital helps.

For example, if Othor AI is testing its AI platform with early customers, this funding can help the team improve accuracy, fix bugs, add features and collect real usage data. That makes the startup stronger when it approaches seed investors later.

Startup aim and purpose

Othor AI’s larger aim appears to be building useful AI technology that can solve real work problems.

That point matters because the AI market is crowded. Many companies now say they are “AI-powered,” but not all of them solve a painful problem. The startups that survive will be the ones that save time, reduce cost, improve decisions or make difficult work easier.

The purpose of Othor AI’s current phase should be simple – build a product that users actually return to. In AI, a flashy demo can get attention, but repeat usage creates a business.

If the startup can show that its tool helps people complete tasks faster or better, the seed round may become easier to raise.

How the funds may be used

Othor AI may use the bridge capital for product improvement, hiring, customer pilots, infrastructure and preparation for its seed round.

Product improvement will be key. AI tools need constant testing because users ask questions in different ways. The system must respond accurately, safely and quickly.

Customer pilots can also help. A pilot is a small trial with a customer. It gives the startup feedback before a full commercial launch.

The company may also spend on cloud and AI model costs. Running AI products can become expensive, especially if many users test the product at once.

Why AI startups are attracting interest

Investors are still interested in AI because the technology is moving into almost every industry. Companies are using AI for sales, support, coding, finance, HR, research, design, legal work and operations.

But investor expectations have changed. Earlier, many startups could raise money only with an AI idea. Now investors want proof.

They ask sharper questions. Does the product have paying users? Is the AI result reliable? Can the startup control costs? Does it have a clear market? Can a big company copy it easily?

For Othor AI, the next stage will likely depend on answering these questions well.

Competitors and market context

Othor AI will operate in a competitive AI market. Depending on its exact product, it may compete with AI productivity startups, automation platforms, enterprise AI tools and global products from companies such as OpenAI, Google, Anthropic, Microsoft and several Indian AI startups.

The challenge for any new AI startup is differentiation. If a product only wraps a common AI model without a strong use case, it becomes easy to copy. But if the startup understands a specific industry problem deeply, it can build a stronger position.

That is why early customer focus is important. A small startup can win if it solves one problem better than larger general tools.

Challenges ahead

• The first challenge is trust. Users need to believe the AI output is useful and accurate.
• The second challenge is cost. AI tools can become expensive to run if the product depends heavily on paid model calls or cloud infrastructure.
• The third challenge is distribution. Even a good product needs customers. Othor AI will need a clear sales plan, whether it targets businesses, professionals or individual users.
• The fourth challenge is seed readiness. The bridge round gives time, but the startup will need traction, product clarity and a strong story before the next fundraise.

Conclusion – Key takeaways

Othor AI’s Rs. 60 lakh bridge funding is an early but useful step before its planned seed round. The capital can help the startup improve its AI product, test the market and build proof for larger investors.

The company’s aim is to build AI-led tools that solve practical work problems. Its next phase will depend on product quality, customer adoption and how clearly it can stand out in a crowded AI market.

For now, the funding shows that even smaller AI startups are finding early support if they can show promise and move quickly.

Facts Input- The Hindu BusinessLine and all other editorial, contextual interpretation, and analytical framing, perspective are added independently sourced from various online mediums/social media/publicly available platforms. Content synthesis and opinion contains original editorial input and all analysis are independently developed & authored.

Emergent has entered the unicorn club after raising $130 million in Series C funding, valuing the AI startup at around $1.5 billion. The company has reached this milestone within a short time of its public launch, helped by the fast rise of “vibe coding” tools.

Emergent was founded by twin brothers Mukund Jha and Madhav Jha. The startup came out of Y Combinator’s 2024 batch and publicly launched in 2025. Its pitch is simple but powerful – let people build software by describing what they want in plain English.

That idea is attracting both users and investors. Emergent reportedly has a revenue run-rate of about $120 million, showing that AI app-building is no longer just a fun experiment for hobby users.

What Emergent does

Emergent is an AI-powered no-code and low-code platform. It helps users build web apps, mobile apps and landing pages without writing traditional code from scratch.

A user can type a prompt such as “build a booking app for a salon with login, payments and appointment reminders,” and the platform starts creating the app. It can ask follow-up questions, generate the structure, connect integrations and help refine the final product.

This is often called vibe coding. The user explains the idea, style and function. The AI handles much of the coding work.

Emergent also offers integrations with tools such as GitHub, Stripe, Supabase and Slack, making it more useful for people who want to build real products instead of only mockups.

Startup aim and purpose

Emergent’s main aim is to make software creation easier for non-coders.

For years, many small business owners, students, creators and professionals had ideas for apps but could not build them because they did not know coding or could not afford a developer team. Emergent wants to reduce that gap.

A shop owner can build a simple inventory tool. A teacher can create a student tracker. A freelancer can make a client portal. A founder can test an app idea before hiring a full engineering team.

The purpose is not only speed. It is access. Emergent is trying to turn software development from a specialist-only skill into something more people can attempt.

Why investors are excited

The funding shows how strongly investors believe in AI-led software development.

Emergent’s growth has been fast. Earlier reports said the company had reached millions of users and crossed major revenue milestones within months. Its valuation has also jumped sharply, from around $300 million earlier this year to about $1.5 billion after the Series C.

That kind of growth tells investors one thing – the market is hungry for tools that make app-building easier.

Companies also like this category because it can serve many types of users. Beginners can use it to build simple tools. Startups can use it to prototype. Developers can use it to move faster. Enterprises can use it for internal apps.

How Emergent may use the funds

Emergent may use the new funding to improve product quality, hire engineering talent, strengthen AI infrastructure, expand enterprise sales and make the platform faster.

Speed will matter. Some reviews have praised Emergent’s guided approach but also noted that it can be slower than some rivals in generating apps.

The company may also invest in better integrations, deployment options, security checks and collaboration tools. These are important if Emergent wants to move from casual builders to serious business users.

Competitors

Emergent operates in a crowded and fast-moving market.

Its main competitors include Lovable, Replit, Base44, Bubble, Hostinger Horizons, Appy Pie and Google’s Opal-style AI app tools. Some focus on professional developers. Some target beginners. Some are stronger in websites, while others focus on full-stack apps.

Emergent’s advantage is its strong early growth and its focus on helping users build more complete apps, not only prototypes. But the category is changing quickly. A feature that feels special today can become common in a few months.

Challenges ahead

The biggest challenge is trust. Building a quick app is impressive, but businesses need apps that are secure, reliable and easy to maintain.

AI-generated code can contain bugs. It may also create security gaps if users do not review it properly. For serious use, platforms like Emergent must offer testing, version control, rollback, documentation and safe deployment.

Another challenge is user expectation. Many beginners may think AI can build everything perfectly in one prompt. In reality, good results still need clear instructions, testing and revisions.

Conclusion – Key takeaways

Emergent’s $130 million Series C funding and $1.5 billion valuation show how quickly AI app-building tools are becoming mainstream.

Founded by Mukund Jha and Madhav Jha, the startup wants to help non-coders build real web and mobile apps using natural language prompts. Its purpose is to make software creation faster, cheaper and more accessible.

The opportunity is huge, but the test is also serious. Emergent must prove that its platform can go beyond demos and support dependable apps for real users and businesses.

Facts Input- ToI

Tata Consultancy Services has launched a new Industrial AI-focused lab in Bengaluru, powered by NVIDIA’s AI infrastructure. The facility, called the TCS Autonomous Engineering Lab Powered by NVIDIA, is located at TCS’ Global Axis campus.

The launch is aimed at one problem many companies are facing right now. They are interested in AI, they run pilots, they test ideas, but moving from a small experiment to a real factory, vehicle platform or shop-floor system is still difficult.

TCS wants this lab to become a practical bridge between AI demos and actual industrial use. The focus is mainly on mobility, manufacturing, autonomous systems, smart plants and engineering-heavy businesses.

What TCS has launched

The new Bengaluru lab will work as a physical hub for Industrial AI innovation. It will help companies design, test, simulate and validate AI-led solutions before using them in real operations.

That last part matters. In manufacturing or mobility, a bad rollout can be expensive. If an AI system fails inside a plant, production may stop. If an autonomous mobility feature is not tested properly, safety becomes a concern. This is why simulation and controlled testing are important.

The lab will use NVIDIA AI infrastructure and TCS’ own industrial engineering experience to help customers build solutions faster and with lower deployment risk.

Why Bengaluru was chosen

Bengaluru is a natural place for such a lab. The city has a deep technology ecosystem, engineering talent, automotive software teams, startups, global capability centres and manufacturing-tech expertise.

For TCS, Bengaluru also gives access to clients, partners and skilled teams that understand both software and industrial engineering. That mix is important because Industrial AI is not only about writing code. It needs knowledge of machines, sensors, vehicles, plants, safety and operations.

What Industrial AI means

Industrial AI simply means using artificial intelligence in physical industries. It is AI for factories, machines, vehicles, supply chains, energy systems and field operations.

A normal office AI tool may summarize emails or write reports. Industrial AI may predict when a machine will fail, inspect product defects through cameras, improve factory output, help vehicles understand roads or simulate how a product will behave before it is built.

This is harder than basic chatbot AI because the system has to work with real-world data, machines, sensors and safety rules.

Key solutions inside the lab

TCS has highlighted several solution areas for the lab.

One is TCS DriveSphere, an AI-led mobility platform for software-defined vehicles. It uses digital twins, real-time data, predictive analytics and over-the-air lifecycle management.

In simple words, it can help vehicle makers manage cars that behave more like connected software platforms. This is becoming important as modern vehicles get more sensors, software updates and driver-assistance features.

The lab will also support AI-led mobility and autonomous systems, including ADAS, autonomous driving, perception systems and intelligent decision-making.

For manufacturing, it will work on physical AI and smart factory use cases such as predictive maintenance, automated quality inspection and real-time process improvement.

The lab will also cover agentic AI and vision AI. Agentic AI can take multi-step actions toward a goal, while vision AI can read images or video from cameras. In a factory, that could mean spotting defects on a production line. In service operations, it could mean helping teams understand problems faster.

Why NVIDIA is important here

NVIDIA is not only a chip company now. Its AI hardware and software stack is used widely for training, simulation, robotics, autonomous driving and industrial AI workloads.

Industrial AI needs heavy computing power. Simulating a vehicle, running vision models or testing a factory process digitally can require strong GPUs and specialized AI tools.

By using NVIDIA’s full-stack AI platform, TCS can offer customers a more serious testing environment. This can help companies try difficult AI use cases before spending money on full deployment.

What companies may gain

For enterprises, the lab can reduce guesswork.

A carmaker can test ADAS-related software in simulation before moving closer to road testing. A manufacturer can test whether vision AI can catch defects on a production line. A plant operator can see if predictive maintenance can actually reduce downtime.

This is useful because many AI projects fail after the pilot stage. They look impressive in a presentation, but struggle when connected to real machines, messy data and business deadlines.

The TCS lab is meant to shorten that distance.

Why this move matters for TCS

TCS is India’s largest IT services company and was founded in 1968 as part of the Tata Group. The company has been trying to deepen its AI services business as clients shift budgets toward automation, cloud, data and AI.

This lab gives TCS a more visible industrial AI asset. Instead of only telling clients that AI can help them, it can bring them into a lab, test use cases and show working prototypes.

That is important because IT services companies are under pressure to prove that AI is not just a consulting story. Clients want results, not slides.

Competitors and market context

TCS will compete with companies such as Infosys, Wipro, HCLTech, Tech Mahindra, LTIMindtree, Accenture, Capgemini and IBM in the industrial AI space.

Several of these firms are also building AI labs, industry clouds, automation platforms and partnerships with chip or cloud companies. The difference will come from execution – who can take AI from experiment to measurable business impact.

For manufacturing and mobility clients, the winners will be partners who understand both software and industrial reality. A factory floor is not a clean test environment. Neither is a road.

Challenges ahead

The biggest challenge is making AI reliable in real operations.

Industrial data is often incomplete, old or scattered across systems. Machines may be from different vendors. Plants may have legacy equipment. Safety checks can slow deployment. Workers also need training to trust and use AI tools.

Another challenge is cost. AI infrastructure is expensive, and companies will want proof that the investment saves money, improves quality or increases output.

TCS will need to show real examples where its lab helps clients move faster and avoid costly mistakes.

Conclusion – Key takeaways

TCS’ NVIDIA-powered Industrial AI lab in Bengaluru is a serious move toward practical AI adoption in mobility and manufacturing.

The lab will help companies prototype, simulate and validate AI solutions before full deployment. Its focus areas include software-defined vehicles, ADAS, autonomous systems, smart manufacturing, predictive maintenance, vision AI and digital twins.

For TCS, the lab strengthens its AI services story. For enterprises, it offers a safer place to test industrial AI before taking it to the real world.

Facts Input- TCS PressRelease

Meta is trying to make a louder entry into the serious AI model race with Muse Spark 1.1, its updated agentic AI model built for coding, tool use and multimodal work.

The timing is interesting. OpenAI, Anthropic and Google have already built strong developer ecosystems around their models. Meta, known for Llama and consumer AI features inside Instagram, WhatsApp and Facebook, now wants developers to take its own in-house model more seriously.

Muse Spark 1.1 is not just another chatbot update. It is being positioned as a model that can help with longer tasks, coding problems, multi-step workflows and inputs like images, videos and documents.

What is Meta Muse Spark 1.1 Agentic Model

Meta Muse Spark 1.1 is the newer version of Meta’s Muse Spark AI model. It is designed for more “agentic” work, which means it can do more than answer a question and stop.

An agentic model can plan steps, use tools, check information, work across apps and continue toward a goal. For example, instead of only explaining a bug in code, it may help identify the problem, suggest a fix and support a larger coding workflow.

Meta has made Muse Spark 1.1 available in Thinking mode through the Meta AI app and website. It is also being opened to developers through the Meta Model API, which is in public preview for developers in the US.

Why this launch matters

Meta has millions of users across WhatsApp, Instagram, Facebook and its smart glasses ecosystem. If it can place a stronger AI model inside those products, AI may become part of daily social, shopping, messaging and work habits.

For developers, the bigger news is API access. Meta is no longer only using its AI inside its own products. It wants outside developers to build apps and tools using Muse Spark 1.1.

That puts Meta in direct competition with OpenAI, Anthropic, Google and xAI, especially in coding and agent workflows.

The pricing angle also matters. Reports say Meta is using aggressive pricing to attract developers. That could put pressure on rivals if the model performs well enough for real production use.

Main capabilities of Muse Spark 1.1

Muse Spark 1.1 is focused on three useful areas – coding, agentic workflows and multimodal understanding.

• In coding, Meta says the model can help detect and fix complex bugs. That is useful because developers do not only need code suggestions. They need help understanding why something broke.
• In agentic work, the model is expected to support end-to-end workflows across apps, including multi-agent systems. Multi-agent systems are setups where more than one AI agent works on different parts of a task.
• In multimodal work, the model can understand images, videos and documents. This can help with tasks like reading a screenshot, checking a file, reviewing a product image or understanding a video-based input.

Coding use cases –

The clearest use case for Muse Spark 1.1 is software development.

A developer could use it to inspect code, explain errors, suggest patches or help with repetitive coding tasks. A startup could connect it to an internal coding tool and use it for bug triage, test writing or documentation.

For example, if a team has a payment bug, the model may help compare logs, read the code path and suggest where the problem could be. A human developer still has to review the final fix, but the model can reduce the time spent searching.

That is the appeal. Good AI coding tools do not replace engineers overnight. They remove some of the slower parts of engineering work.

Multimodal use cases –

Muse Spark 1.1’s ability to understand different input types could make it useful beyond coding.

A product manager may upload a screenshot and ask what is confusing in the user interface. A marketing team may ask it to analyse a video script or a campaign document. A support team may use it to read customer-uploaded images and draft a response.

If Meta connects these abilities deeply with Instagram, WhatsApp and its smart glasses, the model could become more personal and practical. A user wearing smart glasses may ask about something they are seeing. A seller on Instagram may ask AI to create better product replies. A creator may ask for content ideas based on a video.

This is where Meta has an advantage. It already owns the platforms where people communicate, shop, post and create.

Meta Model API and developer push

The Meta Model API is important because it gives developers a way to build with Muse Spark 1.1 outside Meta’s own apps.

Reports say new Meta Model API accounts are getting $20 in free credits. That is a small but useful push to make developers test the model.

If developers like the quality and price, Meta could build a new business line around AI model access. That would be different from its usual ad-heavy business model.

But developer trust takes time. OpenAI, Anthropic and Google already have strong API users. Meta will need reliable performance, clear documentation, stable pricing and good support if it wants developers to stay.

How it compares with rivals

Muse Spark 1.1 will be compared with OpenAI’s GPT models, Anthropic’s Claude models, Google’s Gemini models and xAI’s Grok models.

OpenAI is strong in developer tools and general-purpose AI. Anthropic is known for long-context work, coding and enterprise use. Google has Gemini and deep cloud integration. Meta’s edge may come from lower pricing, social product reach and its ability to place AI inside apps people already use every day.

Still, Meta has to prove the model is not only cheap, but dependable. In business use, a weak answer can cost more than a higher API bill.

Limitations and concerns

There are still open questions.

1. First, availability is limited because the API public preview is for US developers.
2. Second, detailed official documentation and independent testing will matter before companies trust it for critical work.
3. Third, Meta’s AI work often raises privacy questions because of the company’s large social data ecosystem. Users and developers will want clarity on what data is used, how prompts are handled and what control they have.

There is also the safety angle. Meta’s earlier Muse Spark safety report discussed safeguards around high-risk areas such as cybersecurity and chemical or biological misuse. As models become more capable, these controls become more important.

Who should try Muse Spark 1.1

Developers working on coding tools, AI agents, automation products and multimodal apps may find Muse Spark 1.1 worth testing.

Startups with tight budgets may also be interested if Meta’s pricing remains competitive. A cheaper model that performs well can make AI features more affordable at scale.

For normal users, the model may show up through Meta AI, WhatsApp, Instagram or smart glasses features over time. They may not care about the model name, but they will notice if the assistant becomes better at planning, explaining, seeing and helping.

Conclusion – Key takeaways

Meta Muse Spark 1.1 Agentic Model is Meta’s attempt to move from consumer AI features into a more serious developer and enterprise AI race.

Its main strengths are coding support, agentic workflows, multimodal understanding and lower-cost API access. The model could help developers build AI tools, support app-based automation and improve Meta AI across consumer platforms.

The real test will be trust. If Muse Spark 1.1 is fast, affordable and reliable, Meta can become a stronger challenger to OpenAI, Anthropic and Google. If it is only cheap, developers may test it once and move back to models they already trust.

Facts Input- Meta, EvaluationReport