India's AI Cannot Run on Someone Else's Compute — The Billion-Dollar Rise of Neysa

There is a particular kind of founder who does not need to learn the market. They have already built it.

Sharad Sanghi is one of those founders.

He grew up in India, earned his undergraduate degree from IIT Bombay, and then crossed continents to complete his Master's in Electrical Engineering at Columbia University in New York. What followed was not a conventional career in academia or corporate America. Sanghi went deep into the infrastructure of the internet itself — working in the Unified Network Management Architecture Group at AT&T Bell Labs, then in the Backbone Engineering Group at NSFNET, and later in the Router Systems Development Group at IBM Global Network's Advantis division. He helped design and deploy wide area networks for financial institutions including Merrill Lynch and Bankers Trust.

He understood, from the inside, how the digital world's physical foundation was built. And when he returned to India in the late 1990s, he returned with a clarity that very few people in the country's technology sector shared at the time: India needed its own digital backbone. Not someday. Now.

In 1998, he founded Netmagic — India's first commercial internet data centre. The words "data centre" were not yet part of most Indian executives' vocabulary. Sanghi built one anyway.

Twenty-Five Years of Building What Others Called Impossible

Netmagic was not an easy company to build. When the dot-com bubble collapsed in the early 2000s, Sanghi's anchor investor — Exodus Communications — was wiped out. The funding that had underwritten Netmagic's early infrastructure was gone overnight.

Sanghi did not fold. He rebuilt from almost nothing, raised fresh capital from Nexus Venture Partners, Fidelity, Cisco Systems, and Nokia Growth Partners, and turned Netmagic into the dominant data centre and managed hosting provider in India. In 2012, Japan's NTT Communications acquired a 74% majority stake in Netmagic — validating everything Sanghi had built and providing an exit for his venture investors. He stayed on as CEO and Managing Director, continuing to lead the company through its integration into NTT's global infrastructure vision.

In 2016, NTT Data acquired Netmagic fully.

Anindya Das — Sanghi's co-founder at Neysa — had been alongside him through much of this journey, serving as Chief Technology Officer at Netmagic. Together, the two men had spent decades solving exactly the kind of problem that would define the AI era: how to build, scale, and secure large-scale cloud and compute infrastructure in India, under Indian conditions, for Indian and global enterprise customers.

By June 2023, Sharad Sanghi had resigned as Managing Director and CEO of Netmagic. He was ready to build the next version of India's digital backbone. But this time, the backbone was not for websites and enterprise applications.

It was for artificial intelligence.

The Problem That Made Neysa Necessary

In 2023, India's AI ambitions were colliding with an uncomfortable reality.

Companies across India — startups, large enterprises, government organisations — were beginning to understand that generative AI was not a distant future trend. It was an immediate competitive necessity. Every board meeting, as one investor described it, was now a conversation about AI. The demand for AI capability was real, urgent, and growing faster than any reasonable forecast had predicted.

But the infrastructure required to build, train, and run AI models — specifically, high-performance GPU compute — was almost entirely controlled by global hyperscalers: Amazon Web Services, Microsoft Azure, Google Cloud. These were American companies, operating on American terms, with pricing in dollars, data residing outside India, and governance frameworks designed for global markets rather than Indian regulatory and sovereignty requirements.

For Indian enterprises that needed to build AI applications on sensitive data — financial records, healthcare data, government information, personal data of Indian citizens — the hyperscaler model created real, structural problems. Cost was one. Data sovereignty was another. Latency was a third.

Sanghi and Das had spent 25 years building data centre infrastructure in India precisely because they understood that India needed its own digital foundation rather than depending entirely on foreign infrastructure. In 2023, they recognised that the same argument applied, with even greater urgency, to AI compute.

India needed its own AI infrastructure layer. Purpose-built. Sovereign. AI-native from the ground up.

In late 2023, they incorporated Neysa Networks Private Limited in Mumbai and began building.

What Neysa Actually Built

Neysa's mission is to empower builders and businesses with the speed, security, and control they need to innovate confidently in the age of AI.

The company's flagship product is Velocis — an end-to-end AI cloud system that integrates GPU clusters with MLOps toolchains, allowing developers and enterprises to go from idea to production on AI workloads without managing their own hardware. The system is built specifically for machine learning and large language model workloads — not adapted from general-purpose cloud infrastructure, but designed from the ground up for AI.

Neysa operates a GPU-as-a-Service platform called Velocis. The system allows enterprises to rent high-performance computing capacity for training and running artificial intelligence models, rather than building their own hardware.

Beyond GPU-as-a-Service, Neysa offers AI Platform-as-a-Service and Inference-as-a-Service — covering the full spectrum of what an enterprise needs to move from AI experimentation to production deployment. The platform includes dedicated machine learning operations and infrastructure consulting teams who help customers determine the right infrastructure size and customise the models they choose.

Every layer of Neysa's stack — from the cooling systems in its data centres to its orchestration software — is optimised specifically for AI workloads. This AI-native architecture is the fundamental distinction between Neysa and legacy cloud providers who have retrofitted AI services onto infrastructure built for a different era.

Funding That Moved at the Speed of Conviction

The investor community's response to Neysa's founding was immediate and remarkable.

In March 2024 — approximately six months after the company was incorporated — Neysa closed a $20 million seed round led by Matrix Partners India (Z47) and Nexus Venture Partners, with participation from NTTVC. This included a $20 million seed round in 2024 led by Matrix Partners India, Nexus Venture Partners, and NTTVC. It was one of India's largest seed rounds ever raised by a technology startup.

Avnish Bajaj, MD, Matrix Partners India said: "Having known Sharad and Andy for a long time, it is our privilege to partner with them on their next venture. Enterprises globally, and even more so in India, are eager for expertise in helping them transition to AI native cloud computing and there is no better and more experienced team in India for them to partner with."

In October 2024, Neysa raised a $30 million Series A round. Sharad Sanghi formally disclosed the Series A round at the 15th edition of TechSparks 2024, YourStory's flagship conference. The company had by this point launched Velocis to general availability and begun onboarding enterprise customers across key sectors.

Avnish Bajaj, Founder and Managing Director of Z47, emphasised: "Neysa finds itself in the enviable position of having more demand than any other company I have seen at a similar stage and is a catalyst for digital transformation, not just an AI provider."

Then, in February 2026, Neysa announced the largest fundraise in its history and one of the most significant in Indian AI infrastructure: a $600 million equity round led by Blackstone, accompanied by a $600 million debt facility — a total capital deployment of $1.2 billion to finance the expansion of Neysa's GPU fleet to 20,000-plus units across its Mumbai and Hyderabad data centres.

Neysa has raised $650 million in funding from investors including Z47, Nexus Venture Partners, and Blackstone, with a current valuation of $1.4 billion.

Neysa had become a unicorn in under three years from founding — one of the fastest unicorn trajectories in Indian startup history.

The Marketing Strategy Built on Institutional Trust and Proven Pedigree

Neysa's marketing strategy is unlike almost any other Indian startup of its era — because it is built not on virality, community, or consumer awareness, but on three deeply specific forms of institutional credibility.

The founders as the brand. Neysa did not launch as an unknown entity trying to earn trust from scratch. It launched with the accumulated credibility of two people who had spent 25 years building the most significant data centre infrastructure company in India. Every enterprise that knew Netmagic — and virtually every large Indian organisation did — already had a reference point for what Sanghi and Das could build. The founding team's track record was not a background detail. It was the primary marketing asset.

Sovereign AI as a national narrative. At a moment when India's government and corporate establishment were deeply engaged in conversations about data localisation, AI sovereignty, and reducing dependence on foreign technology infrastructure, Neysa's positioning as India's purpose-built AI infrastructure platform was precisely aligned with the dominant policy and strategic conversation. The company did not have to manufacture relevance. The relevance of what it was building was self-evident to every enterprise that had ever worried about where its AI data lived.

Institutional investors as credibility amplifiers. The $20 million seed round — one of India's largest ever — led by Matrix Partners India and Nexus Venture Partners, was not just capital. It was a public signal, from two of India's most respected technology venture firms, that Neysa represented the highest-conviction infrastructure bet available. The subsequent Blackstone-led Series B — from one of the world's largest and most disciplined private equity firms — extended that signal to the global institutional market.

Demand that exceeded supply as the proof point. Avnish Bajaj's publicly stated observation — that Neysa had "more demand than any other company I have seen at a similar stage" — is not a marketing claim made by the company. It is an independent observation from a board member and institutional investor with decades of experience evaluating Indian startups. This kind of third-party validation, from a credible source with no incentive to overstate, is more powerful than any campaign a marketing department could design.

Product visibility through conference positioning. Neysa chose TechSparks 2024 — YourStory's flagship technology conference, one of India's most prominent platforms for tech entrepreneurship — as the venue for its Series A announcement. This was a deliberate choice to maximise visibility among the Indian technology community at a single, high-profile moment rather than spreading announcements across multiple smaller channels.

A Clean Slate for India's AI Future

The name Neysa, in several cultural traditions, means pure or intelligent. For Sanghi and Das, it represented something more specific: a clean slate. An infrastructure platform free from the legacy architecture of 20-year-old cloud systems, designed from the ground up for a workload that did not exist when most cloud infrastructure was built.

India's AI era is not a distant possibility. It is the present moment. The enterprises building AI applications, the developers training models, the government organisations deploying AI at scale — all of them need compute infrastructure that is fast, sovereign, cost-effective, and built by people who know how to deliver at Indian scale.

Neysa's founders built India's first commercial data centre in 1998 when nobody understood what that meant. They scaled it into the country's dominant infrastructure provider over 25 years. They sold it to NTT. And then they started again.

Because they saw, more clearly than almost anyone else in India, that the next layer of the country's digital foundation needed to be built. And that nobody was better positioned to build it.

From IIT Bombay to AT&T Bell Labs to Netmagic to Neysa — Sharad Sanghi has spent his career building the infrastructure that India runs on.

He is doing it again. And this time, it is for the age of AI.

Founded 2023. Mumbai. $20M seed in six months. $1.4B valuation in three years. Building India's AI backbone.