AI to move from pilots to production, see wider adoption in 2026: Neysa’s Sharad Sanghi

Artificial intelligence (AI) will see wider adoption by enterprises as the technology moves beyond proofs-of-concept to full-scale production, said Sharad Sanghi, chief executive officer, Neysa. Overall sharpening of the models will reduce hallucinations and make them more reliable for business adoption, he added.

Sanghi, who till 2023 was the chief executive officer of Japanese IT services major NTT in India, said the industry is close to achieving artificial general intelligence (AGI) around 2027.

Neysa is a cloud-based AI acceleration platform that offers compute power and software tools to build, run, and manage AI applications for businesses and organisations large and small. The Mumbai-based company was founded by Sanghi and former senior Netmagic executive Anindya Das, and is backed by the likes of NTTVC, Z47, Nexus Venture Partners, the Anchorage Capital Group, etc.


The Mumbai-based company last raised $30 million in 2024 (about Rs 252 crore) in a round co-led by the US-based venture capital firm NTTVC, Z47, and Nexus Venture Partners, which valued the it at $76 million.

Citing a recent MIT study which found that 95% of agentic AI deployments are yet to move to production, Sanghi said that that will change over the next 12 months, marking a major shift in enterprise adoption this year.

ET had reported in December that venture capital firm Lightspeed had entered the fray to pick up stake in the cloud infra startup, joining Japanese investor SoftBank and private equity firm Blackstone, which had also been in talks. This could take the Neysa’s valuation to as much as $250-$300 million, sources had told ET then.

Sanghi explained that companies are adopting AI primarily for two purposes: improving productivity, and competitive differentiation. The growing adoption of tools such as copilots and AI-based coding assistants, Sanghi said, is helping enterprises enhance developer productivity. “Broadly, we see two key drivers in the market — using AI to reduce costs through increased efficiency, and using it to innovate and expand business opportunities.”

He added that frontier AI labs are making steady progress and the technology may surpass human capabilities in almost every area soon. But challenges around explainability, ethics, and fairness remain. “Issues are still there, but model quality and accuracy are improving rapidly, and hallucinations are becoming less frequent,” Sanghi noted. “We’re steadily moving closer to what can be described as AGI.”

The India chapter

Neysa positions itself at the intersection of cloud computing and AI infrastructure, two fields drawing the most investor interest. Global hyperscalers and Indian conglomerates are expected to invest more than $50 billion in the next five to seven years in the domestic data centre industry, taking the total capacity to about 9 gigawatts (GW) from 1 GW now.

Tech giants such as Google, Amazon Web Services, Microsoft Azure, and Meta are making big bets in the AI infra space, along with homegrown players such as Adani, Bharti Airtel, Reliance Industries, TCS, and L&T. Given their massive capex, hyperscalers are leading. Analysts estimate over $500 billion will be spent globally in 2026 alone on AI infra.

A key example is OpenAI's ambitious Stargate Project, a $500-billion initiative announced in early 2025 to build gigawatt-scale data centres in the US, with partners like Nvidia, Oracle, SoftBank, and Microsoft.

Since then, Google has also upped its game. The tech giant, which recently crossed $4 trillion in market cap for the first time, announced a $15 billion investment in October 2025, to establish its first AI hub and the largest data centre campus outside the US in Andhra Pradesh's Visakhapatnam.

Microsoft plans to invest $3 billion in India over the next two years in cloud and AI infra, including the establishment of new data centres.

Mukesh Ambani's Reliance Industries formed a new wholly-owned subsidiary called Reliance Intelligence for large-scale AI infrastructure in the country. IT major TCS plans to invest up to $7 billion in a 1 gigawatt data centre.