India’s scale, linguistic diversity and growing focus on AI sovereignty will make it one of the most important markets for smaller, cost-efficient artificial intelligence models, Databricks chief executive Ali Ghodsi has said, as the company unveiled a new suite of AI agent technologies at its Data + AI Summit.
Speaking to reporters, Ghodsi said Databricks is working with local partners in India, including Reliance, as enterprises and government organisations seek sovereign AI and cloud infrastructure.
Databricks ties up with Reliance on AI expansion
“There is a whole region that is being developed by Reliance with local GPUs and local infrastructure, and we are active in that,” Ghodsi said.
Reliance has separately committed massive investments in AI infrastructure, including plans to invest roughly $110 billion to position India as an AI hub, building AI-ready infrastructure through Jio.
Ghodsi added that India’s price-sensitive market and hundreds of languages create demand for smaller open-source and regional AI models alongside frontier proprietary systems.
Globally, concerns against restrictions on open-source AI models have been flagged. These models remain critical for research and innovation despite proprietary frontier models continuing to lead on performance.
Databricks launched Genie One, its new AI-powered agentic coworker designed to help businesses automate tasks and make decisions using enterprise data. The platform targets functions such as finance, sales, marketing and operations, allowing employees to ask questions, generate insights and execute tasks using natural language.
The launch places Databricks more directly in competition with enterprise AI offerings from Microsoft, Salesforce, Google Cloud and OpenAI, all of which are racing to develop AI agents capable of automating workplace tasks.
The San Francisco-based company, which is valued at $134 billion, has also introduced Genie Agents, which can automate actions across enterprise systems, and Genie Code, a tool focused on data engineering and machine learning workloads.
The company used the summit to acknowledge concerns around AI spending among customers. Ghodsi said that the conversation has now shifted from AI adoption to AI economics, with focus on controlling costs while expanding AI usage.
Unity AI Gateway routes workloads across models
To address that challenge, Databricks launched Unity AI Gateway, which automatically routes workloads to different models based on task complexity. Simpler requests can be handled by smaller or open-source models, while more advanced workloads are directed to frontier systems.
Reliance Industries Chairman and Managing Director Mukesh Ambani said virtually that the group is leveraging the Databricks Data Intelligence Platform to unify and govern petabytes of data across its diverse businesses. The centralized data architecture supports Reliance’s AI ambitions by connecting data from Jio’s subscribers, retail customers, petrochemicals operations, and emerging new energy businesses.
Databricks also plans to hire more than 600 graduates this year, more than doubling previous intake levels, as it expands its AI and data platform business globally.
The company said its AI business is now generating an annualised revenue run rate of more than $1.7 billion, up from more than $1 billion a year ago.
(The writer was in San Francisco at the invitation of Databricks)