Agnikul Cosmos, a private Indian space transportation company, has announced a commercial partnership with NeevCloud to deploy AI-powered data center modules in Low Earth Orbit (LEO). The collaboration marks one of the first initiatives globally to host AI inference workloads on a convertible rocket upper stage in space.
Under the agreement, Agnikul will launch NeevCloud’s orbital AI data center module using its patented extendable upper-stage platform. The system will enable real-time AI inference directly from orbit, eliminating the need for customers to independently design or deploy satellites.
NeevCloud becomes one of the earliest enterprises to test AI inference and cloud workloads on Agnikul’s convertible upper-stage platform. The approach opens a new revenue stream for Agnikul through integrated launch and in-orbit hosting services, while offering enterprises a commercially viable pathway to access space-based compute infrastructure.
By placing compute capabilities in orbit, the partnership addresses latency limitations associated with terrestrial data centers. The space-based model enables real-time AI inferencing for regions with limited or unreliable ground infrastructure, expanding AI accessibility across underserved geographies.
Agnikul plans to deploy a fleet of modular Space Data Center Modules (SDCMs) equipped with AI stacks. NeevCloud intends to scale its orbital network to more than 600 edge data centers over the next three years, creating a distributed constellation capable of offloading intensive computing workloads from terrestrial devices to AI processors in space.
Srinath Ravichandran, Co-founder and CEO of Agnikul Cosmos, stated that the company’s convertible upper-stage technology allows rocket stages traditionally left in orbit to remain operational and host computing hardware and software. This extends the lifecycle and utility of launch infrastructure beyond payload delivery.
Narendra Sen, Founder and CEO of NeevCloud, described the initiative as the development of an orbital inference layer designed to expand AI access beyond terrestrial constraints. Moin SPM, Co-founder and COO of Agnikul Cosmos, added that lower upfront launch costs and flexible service models make space infrastructure increasingly accessible to enterprises.
The first pilot launch is scheduled before the end of the year, with broader deployment planned following validation. The collaboration positions India at the intersection of the rapidly expanding AI inference market and the emerging space-compute economy, while advancing sovereign, cost-efficient, and sustainable orbital infrastructure.
NeevCloud AI Supercloud is making high-performance computing more accessible, affordable, and sovereign. It delivers a full-stack, end-to-end AI infrastructure that empowers enterprises and governments with private AI capabilities and complete control over their data and compute. Built to support mission-critical applications across key sectors—including BFSI, healthcare, automotive, startups, and research—NeevCloud’s sovereign infrastructure helps drive national innovation. With a vision to democratize AI for the next billion users, NeevCloud offers compute costs 40–60% lower than global hyperscalers, strengthening India’s position as a global leader in the AI revolution.
Agnikul Cosmos Private Limited, an IIT Madras–incubated space transportation company based in Chennai, is developing on-demand launch vehicles called Agnibaan to carry small satellites into orbit. The company achieved a major milestone last year with its maiden launch from India’s first private launchpad. This mission featured a 6-axis controlled ascent flight powered by a single-piece 3D-printed engine designed and manufactured entirely in-house, along with proprietary autopilot algorithms that successfully guided the vehicle along its planned trajectory.
