New Delhi, Nov. 5 -- Indian enterprises are moving beyond generative AI experimentation to deploy the technology across core business functions, but fragmented data and the lack of business context remain significant barriers to scaling adoption, according to U.S-based data cloud company Snowflake.

India is emerging as one of the leading markets for generative AI adoption globally, with 71% of Indian respondents reporting measurable returns from their investments, compared with 61% globally, according to a Snowflake survey conducted with Omdia. The study also found that 42% of Indian organisations are already using generative AI across multiple use cases.

"Organisations are now moving from AI experimentation to operationalising AI, with production use cases emerging across industries," said Vijayant Rai, managing director, India, Snowflake.

The adoption is particularly advanced among digital-native companies, while traditional enterprises are increasingly deploying AI across IT operations, analytics, cybersecurity, software development, customer service and supply-chain functions. Snowflake's survey found that 76% of Indian respondents reported generative AI adoption in IT operations, against 61% globally, while adoption in data analytics stood at 75%, compared with 57% globally.

However, scaling these applications requires more than access to large language models (LLMs). Rai identified data silos and the absence of business-specific context as the two biggest challenges facing traditional enterprises.

AI systems need to understand an organisation's definitions, processes and relationships to generate reliable outcomes. For instance, the meaning of terms such as revenue or customer can differ across businesses, making contextual accuracy critical as companies move towards autonomous AI agents.

"The two biggest challenges we see are data and context. Model choice comes next," Rai said, adding that enterprises also need to balance model performance with token costs and returns on AI investments.

Snowflake is positioning its platform as an agentic control plane that brings together enterprise data, AI models, governance and application development. Its approach allows customers to access multiple AI models, including those from OpenAI, Anthropic, DeepSeek and Mistral, while keeping enterprise data within its governed environment.

The company is also seeing a larger role for system integrators (SIs) and technology partners as enterprises seek industry-specific AI solutions. Dhiraj Narang, director and head of partnerships, India, Snowflake, said the company's partner attach rate is close to 100%, with partners involved in virtually every field opportunity.

The partner ecosystem includes global system integrators such as TCS, Infosys, LTIMindtree, Persistent and Coforge, alongside advisory and technology partners. Narang said the focus is shifting towards enabling partners to combine domain expertise with AI capabilities to develop solutions faster, particularly in areas such as supply-chain management, procurement and demand forecasting.

At Snowflake Expedition 2026 on 4 November, the company announced new capabilities covering business definitions, AI governance, model routing and autonomous workflows. These include Business Ontology and Cortex AI Gateway within Snowflake Horizon Catalogue, alongside enhancements to its CoCo coding agent and CoWork work agent. The updates are aimed at helping enterprises establish business context, manage AI costs and govern agent-driven workflows.

For India, the company is pursuing a three-pronged strategy: expanding its domestic customer base, building capabilities and partnerships in India for global markets, and encouraging technology partners to develop solutions on its platform.

Rai said India was a key growth engine for the company, with its partner ecosystem in the Asia-Pacific and Japan region increasingly supporting customers beyond the domestic market.

Published by HT Digital Content Services with permission from MINT CIO CIRCLE.