Databricks at a crossroads: Can its AI strategy prevail without Naveen Rao?

Databricks finds itself in an awkward situation following the departure of Naveen Rao, its head of artificial intelligence, as rivals like Snowflake, Teradata, and hyperscalers such as AWS, Azure, and Google Cloud, intensify their push to develop offerings for building generative AI applications.

Rao’s exit comes at a time when Databricks is aggressively trying to expand its offerings inside the Data Intelligence Platform, mainly with Lakebase and Agent Bricks, thanks to the infusion of capital from this week’s $1 billion Series K funding round that saw its valuation surge past $100 billion.

As a result, Rao’s departure introduces uncertainty and raises questions about the company’s ability to sustain innovation velocity and technical leadership in a fiercely competitive market.

“Databricks is in a tricky spot with Naveen Rao stepping back. He was not just a figurehead, but deeply involved in shaping their AI vision, particularly after MosaicML,” said Robert Kramer, principal analyst at Moor Insights & Strategy.

“Rao’s absence may slow the pace of new innovation slightly, at least until leadership stabilizes. Internal teams can keep projects on track, but vision-driven leaps, like identifying the ‘next MosaicML’, may be harder without someone like Rao at the helm,” Kramer added.

Rao became a part of Databricks in 2023 after the data lakehouse provider acquired MosaicML, a company Rao co-founded, for $1.3 billion. During his tenure, Rao was instrumental in leading research for many Databricks products, including Dolly, DBRX, and Agent Bricks.

That leadership gap, the analyst warned, could give its rivals a convenient narrative to draw away customer attention.

“To counter any such narrative, Databricks needs to prove quickly that AI remains central to its Data Intelligence Platform and not just a layer on top. If they falter, rivals, specifically, hyperscalers, will use their speed and bundled services to lure customers,” Kramer said.

And rivals such as Snowflake are not sitting idle either. While Snowflake has continued to add features that rival most of Databricks’ AI and ML offerings, others such as Teradata and Cloudera are using strategies such as repositioning with AI factories and leaning into open standards such as Iceberg and the Model Context Protocol (MCP) to strengthen their AI approaches.

However, The Futurum Group’s lead for data intelligence and analytics, Bradley Shimmin, unlike Kramer, doesn’t foresee trouble for Databricks after Rao’s departure.

The AI market has shifted from building frontier models to integrating them, and technology vendors now compete by how well they apply these models, not by owning them, Shimmin said, pointing out that MosaicML’s contribution is passé.

Further, the analyst believes that Databricks will not lose ground to rivals like Snowflake or Teradata, as “the company already enjoys a comfortable lead, technologically speaking, in both building and running AI.”

No immediate successor

While analysts remain divided over the impact of Rao’s departure, the company is yet to name his successor despite the role of head of AI being central to its future product roadmap.

An email sent to Databricks specifically asking about succession did not elicit a response.

However, Kramer said that the company is more likely to rely on internal leaders who already know the platform, rather than rushing into an external hire.

“This approach helps maintain continuity but also risks leaving a gap in outward-facing thought leadership. External recruitment could happen down the line, but in the short term, it looks like they’ll tap existing product and research teams to carry forward Rao’s priorities,” Kramer said.

Offering a different view, Shimmin said that Databricks might not immediately feel the vacuum left by Rao, and can rely on CEO Ali Ghodsi and CTO Matei Zaharia.

After all, he said, “you’ve got two technology-first executives with steady hands on the steering wheel. With Matei in particular, you’ve got someone who literally created and open-sourced much of the technology upon which Databricks is built, not just Apache Spark, but also MLflow, and Delta Lake.”

Another fork in the road for Databricks?

Rao’s departure, according to Kramer, also presents Databricks with a critical choice: focus more on execution, making steady progress on the AI capabilities already underway, or chase the next big bet.

For Databricks, the next big bet could be innovating on balancing the cost and efficiency of its offerings, Kramer said, adding that one way to achieve that would be specialized hardware for AI.

Targeted hardware for AI can bring down the cost of training models or running complex queries to generate insights.

Databricks is also investing in Rao’s startup, CEO Ali Ghodsi wrote on LinkedIn, and confirmed that, as Rao had already hinted in a post, it would focus on the AI hardware space.

Rao, when asked directly about his startup, said that he would provide more details next week.

For now, especially after Rao’s departure, Databricks may have to lean more on partnerships, although, given the company’s history of acquisitions and all the tell-tale signs, Rao and his new startup may in the future become part of the company again.