Fresh off a $1.2 billion funding round and the unicorn crown, Neysa is preparing to accelerate its growth across enterprise AI, GPU infrastructure, and data centre partnerships.
In an exclusive interaction with AIM, Karan Kirpalani, Chief Product Officer at Neysa, said that the priority is scaling rapidly to meet growing demand.
“For us, it’s to be able to fuel our growth. We’ve been growing pretty rapidly,” Kirpalani says. “We’ve made so much headway over the last two or three quarters, specifically into the enterprise segment, including medium and large enterprises across verticals.”
He adds that Neysa is seeing strong traction among AI-native independent software vendors, fintech firms, SaaS companies, model creators, and developers building industry-focused AI use cases.
The Mumbai-based company operates at the infrastructure layer that underpins AI systems rather than building end-user AI products. Through its Velocis platform, the company offers GPU-as-a-service, giving businesses on-demand access to high-performance compute capacity. It allows customers to train, fine-tune, and deploy AI workloads.
Neysa was founded in 2023 by Sharad Sanghi and Anindya Das. Sanghi previously founded data centre services provider Netmagic Solutions, which was acquired by NTT Communications in 2012, bringing deep infrastructure expertise to Neysa.
GPUs Loading
Neysa plans to bring in 20,000 GPUs. According to Kirpalani, the company has already started placing orders. “We’re almost at capacity in terms of our existing cohort of GPUs. We’re starting to see requirements for very large cohorts from single customers.”
The company expects NVIDIA’s latest Blackwell GPUs, including B300s, to form a major part of the expansion.“Our Blackwells are already on the way. The IndiaAI Mission is a source of utilisation for us as well. We need those GPUs as quickly as possible to meet a lot of those requirements.”
Kirpalani said one of the IITs has placed an order for more than 200 servers. He added that several AI-native startups, including Shodh AI, Navana.ai, Smallest.ai, GreyLabs AI, Protecto, and Kogo AI, are building and deploying their models on Neysa’s Velocis Cloud.
Neysa currently operates about 1,200 GPUs. In 2024, it raised $50 million, split between two rounds.
Besides Neysa, Yotta also recently announced that it will deploy 20,736 liquid-cooled NVIDIA Blackwell Ultra GPUs to form its AI superclusters. Similarly, NxtGen has also partnered with Dell Technologies to build an AI factory with 4,000 NVIDIA Blackwell GPUs.
Neo Cloud vs Hyperscalers
Kirpalani distinguishes between general-purpose hyperscalers such as Amazon, Microsoft Azure, and Google Cloud, and purpose-built AI neo clouds like Neysa.
“We are architected from the ground up to allow clients to do training, fine-tuning, and inferencing in the most efficient and cost-effective way,” Kirpalani notes.
He adds that this design translates into significant cost savings. “We’ve been able to migrate some clients away from traditional hyperscale clouds by showing more than a 70–80% reduction in total cost of ownership.”
In one case, Kirpalani says Neysa was able to deliver nearly 90% cost savings compared with proprietary models running on hyperscale clouds. He adds that the company improves “time to first token and token throughput,” while also offering committed quality of service on those metrics.
Data Centre Partnerships
Neysa’s GPU infrastructure is currently hosted on NTT data centres in India, though the company is evaluating other providers as it scales.
“We want to work very closely with our data centre partners, not just consume services, but co-create and engineer with them to ensure maximum efficiency in cooling and power density,” Kirpalani notes.
He emphasises the importance of managing the full AI supply chain, including compute, networking, and storage partners. “As long as we make sure that all elements of our supply chain are uninterrupted, we can continue scaling.”
Despite broader industry concerns around NAND and DRAM flash memory shortages, Kirpalani said Neysa has not faced significant disruptions so far. Kirpalani says that whenever Neysa procures NVIDIA GPUs through original equipment manufacturer (OEM) partners, NVIDIA is part of those conversations, ensuring the company has clear visibility into its supply chain.
Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.