HCLTech and Sarvam AI Announce Rs 14,257 Crore Odisha AI Data Centre on 24 July 2026
HCLTech announced on 24 July 2026 that it will build its first dedicated AI data centre in a three-way partnership with Sarvam AI and the Government of Odisha. The planned capital outlay is Rs 14,257 crore — approximately $1.4 billion at current exchange rates — with financial assistance from the state government forming part of the investment structure. The announcement was made at the Utkarsh Odisha investment summit in Bhubaneswar. This is the first purpose-built AI data centre that HCLTech, one of India's three largest IT services companies by revenue, has committed to building. The company has historically operated technology delivery centres and co-located infrastructure rather than investing in dedicated AI compute facilities. The Odisha commitment marks a structural shift in how established Indian IT services leaders are positioning for the AI infrastructure cycle.
Why Odisha, and What the Sovereign AI Park Offers
The data centre will be built in the Odisha Sovereign AI Park in Bhubaneswar — a state government initiative to attract large-scale AI infrastructure investment by offering land, power connectivity, regulatory facilitation, and direct financial participation. India's central government has been encouraging states to develop dedicated AI infrastructure zones as part of the IndiaAI Mission, which aims to build sovereign compute capability domestically rather than relying entirely on US-hosted cloud platforms. The Odisha Sovereign AI Park is designed as an anchor-tenant facility: HCLTech's Rs 14,257 crore commitment is the first and largest investment, intended to attract further co-tenants from AI infrastructure, research, and application development sectors. The choice of Bhubaneswar signals that India's AI infrastructure investment is no longer geographically constrained to the four established technology corridors of Bengaluru, Hyderabad, Chennai, and the National Capital Region. State capitals with strong engineering university ecosystems, competitive land costs, and sufficient power grid capacity are now viable locations for large-scale AI compute.
Sarvam AI's Role: India's Open-Source Foundation Models
Sarvam AI is India's most prominent open-source foundation model company. Founded to serve India's linguistic diversity, it has built a series of open-weight multilingual models trained specifically on Indian language data — covering Hindi, Tamil, Telugu, Kannada, Marathi, Bengali, Gujarati, and other Indian languages. Its models have been integrated into government digital public infrastructure and enterprise applications where Indian-language capability is a core requirement that general-purpose global models do not consistently meet. In the HCLTech partnership, Sarvam AI's role is to provide the foundation models that the Odisha facility will host and serve at scale. HCLTech will contribute systems integration, cloud architecture, and enterprise application delivery. Together they intend to build and serve sector-specific AI services for healthcare, agriculture, finance, public services, and education — domains where Indian-language fluency and domain-specific fine-tuning are essential and where offshore model providers operating on English-centric training data fall measurably short.
Scale and Timeline: 5,000 Employees and a 2028 Launch
The data centre will house 5,000 employees when operational, making it a significant employment anchor for Bhubaneswar. Operations are targeted to begin in 2028. That timeline is consistent with the construction and infrastructure commissioning lead time for a facility at this scale: large AI data centres require approximately 18 to 36 months from groundbreaking to first operational racks, given the power, cooling, and high-density networking infrastructure required. The project also includes a separate HCLTech technology centre co-located within the Sovereign AI Park, serving as the primary employee campus adjacent to the compute facility. The combined announcement positions Odisha as the site of one of the largest single-site AI infrastructure investments ever made by an Indian company on Indian soil.
What This Investment Means for the Indian AI Ecosystem
The HCLTech-Sarvam-Odisha announcement matters beyond its headline figure. India's two most visible gaps in the global AI competition have been the absence of a domestically hosted, India-governed GPU cluster at commercial scale, and the absence of foundational models trained on the full breadth of Indian linguistic and cultural data. This project directly addresses both. When operational in 2028, the Odisha facility will add substantial AI compute capacity that is India-hosted, India-governed, and built around Indian-language foundation models from Sarvam. For regulated industries — banking, financial services, insurance, healthcare, and public sector — this will provide a credible data-residency-compliant alternative to US-hosted cloud AI services. For Indian-language AI product companies, it will lower the barrier to serving the hundreds of millions of Indian users who consume digital services in languages other than English.
What This Means for Indian Software and AI Product Teams
For Indian software companies and AI product studios, the practical implication of the HCLTech-Sarvam partnership is a clearer roadmap for when Indian-language foundation model infrastructure will be available at cloud scale through an established enterprise partner. Teams building products for Tier 2 and Tier 3 India — where Hindi, Tamil, Telugu, Bengali, and other regional languages dominate — can plan for a 2028 horizon at which domestically hosted model inference will be commercially accessible. Teams serving regulated sectors where data residency is a compliance requirement should incorporate this timeline into their architecture decisions, particularly for workloads currently routing through US-hosted model providers. The partnership also validates Sarvam AI's model development path: with HCLTech and state government infrastructure behind it, Sarvam's Indian-language models are on course to reach production scale through institutional distribution channels, not just developer API usage.
The Bottom Line
On 24 July 2026, HCLTech, Sarvam AI, and the Government of Odisha announced a Rs 14,257 crore (approximately $1.4 billion) AI data centre investment in Bhubaneswar's Sovereign AI Park — HCLTech's first purpose-built AI compute facility and one of the largest single-site AI infrastructure investments ever made by an Indian company. The facility will use HCLTech's systems integration capabilities and Sarvam AI's Indian-language foundation models to serve sector-specific AI applications for public and private enterprises across healthcare, agriculture, finance, and public services. It will house 5,000 employees and target a 2028 start of operations. For Indian software and AI product teams, the project's most significant long-term implication is a domestic, data-residency-compliant AI infrastructure option that natively covers Indian languages — closing a capability gap that has persistently constrained what Indian-language AI products can practically build and deploy at scale.
Frequently Asked Questions
What is HCLTech and Sarvam AI's announced AI data centre project in Odisha?+
On 24 July 2026, HCLTech and Sarvam AI announced a three-way partnership with the Government of Odisha to build a dedicated AI data centre in the Odisha Sovereign AI Park in Bhubaneswar. The planned capital outlay is Rs 14,257 crore — approximately $1.4 billion — with financial assistance from the Government of Odisha as part of the investment structure. The facility is HCLTech's first purpose-built AI data centre. It will house 5,000 employees, target operations in 2028, and use HCLTech's full-stack AI integration capabilities alongside Sarvam AI's Indian-language open-weight foundation models to serve sector-specific AI applications for private and public sector enterprises.
What is Sarvam AI and why is it central to the Odisha data centre project?+
Sarvam AI is India's most prominent open-source foundation model company, built specifically to serve India's linguistic diversity. It has developed open-weight multilingual models trained on Indian language data covering Hindi, Tamil, Telugu, Kannada, Marathi, Bengali, Gujarati, and other Indian languages. Its models have been integrated into government digital public infrastructure and enterprise applications where Indian-language capability is essential. In the HCLTech-Odisha data centre partnership, Sarvam AI provides the foundation models the facility will host and serve at scale. This means the data centre's AI output will be built around Indian-language models rather than general-purpose global models that were not trained on Indian linguistic and cultural data — a meaningful distinction for enterprises and public sector organisations serving large Indian-language user bases.
What is the Odisha Sovereign AI Park and how does it fit into India's AI policy?+
The Odisha Sovereign AI Park is a state government initiative in Bhubaneswar designed to attract large-scale AI infrastructure investment by providing land, power connectivity, regulatory facilitation, and direct financial participation. It is part of India's broader IndiaAI Mission, which aims to build sovereign compute capability domestically rather than relying entirely on US-hosted cloud platforms. Odisha's Sovereign AI Park is designed as an anchor-tenant facility: HCLTech's Rs 14,257 crore commitment is the first and largest investment, expected to attract further co-tenants from AI infrastructure, research, and application sectors. The park positions Odisha as one of the first states to anchor a major AI infrastructure commitment outside the four established technology corridors of Bengaluru, Hyderabad, Chennai, and the National Capital Region.
When will the HCLTech AI data centre in Odisha become operational and who will benefit?+
HCLTech and Sarvam AI have targeted 2028 as the start of operations for the Odisha AI data centre, consistent with the 18-to-36-month lead time that large AI data centres require from groundbreaking to operational racks. The facility will primarily benefit three categories of users when operational: Indian enterprises in regulated sectors such as banking, financial services, insurance, and healthcare that require data-residency-compliant AI infrastructure; public sector organisations and government agencies seeking India-hosted AI compute for sensitive applications; and Indian-language AI product companies building applications in Hindi, Tamil, Telugu, and other regional languages who need foundation model inference served from India at commercial scale.
Written by
TechPillow Team
Sharing insights on technology, product development, and the Indian tech ecosystem.