Databricks CEO says SaaS isn’t dead, but AI will soon make it irrelevant

On Monday, Databricks disclosed that it has reached a $5.4 billion revenue run rate, marking 65% year-over-year growth. Of that amount, more than $1.4 billion was generated from its artificial intelligence product offerings.

Co-founder and CEO Ali Ghodsi shared updated growth metrics amid growing debate over whether AI could disrupt or even replace the traditional SaaS business model. According to Ghodsi, AI has not weakened Databricks’ position. Instead, it has driven increased adoption and deeper usage of its products.

He acknowledged that many in the industry are questioning the future of SaaS as AI continues to evolve. However, in Databricks’ case, AI has functioned as a growth catalyst rather than a destabilising force.

At the same time, Ghodsi has avoided characterising Databricks purely as a SaaS company, particularly since private-market investors now tend to value it as an AI-driven enterprise. Alongside its revenue update, the company confirmed that it has officially closed its previously announced $5 billion fundraising round at a $134 billion valuation and secured an additional $2 billion loan facility.

Even with its AI positioning, Databricks remains rooted in the cloud data warehouse market, which formed the foundation of its business. Cloud data warehouses enable organisations to store and analyse large datasets to extract business intelligence and operational insights.

Ghodsi highlighted a specific AI-powered product contributing to increased engagement: Genie, the company’s large language model-based interface. Genie illustrates how a conventional SaaS platform can be transformed by replacing its traditional user interface with a conversational, natural-language layer.

For instance, Ghodsi explained that he uses Genie to understand why data warehouse usage and revenue may spike on certain days. Previously, answering such questions would have required writing technical queries in specialised languages or requesting a customised report. Now, with an LLM-driven interface, those inquiries can be handled conversationally, broadening access to complex data analysis.

He indicated that Genie has played a notable role in driving the company’s growth metrics by reducing friction in user interactions with sophisticated data systems.

Ghodsi also challenged the idea that AI will prompt enterprises to abandon their established SaaS systems of record in favour of internally developed, AI-generated alternatives. Systems of record manage essential business information across areas such as sales, customer service, and finance, and replacing them is often complicated and risky.

In his view, large AI model providers are not attempting to replace enterprise databases or core infrastructure. Rather, they aim to modernise how users access and interact with those systems—through natural language interfaces for human users or through APIs and integrations for AI agents.

He suggested that the most immediate impact of AI on SaaS companies may be cultural rather than structural. Historically, professionals built careers around mastering specific enterprise platforms such as Salesforce, ServiceNow, or SAP. As interfaces shift toward natural language, those specialised systems may recede into the background.

Ghodsi noted that millions of professionals worldwide have been trained on specific SaaS user interfaces, which historically created a strong competitive moat. As interactions become language-based, the distinctiveness of those interfaces may diminish, making products less visible and more akin to underlying infrastructure.

SaaS providers that embrace LLM-powered interfaces could continue to expand, as Databricks has done. However, the transition also creates opportunities for AI-native companies to introduce products designed from inception to integrate seamlessly with AI tools and autonomous agents.

To position itself within this evolving environment, Databricks introduced Lakebase, a database tailored specifically for AI agents. Ghodsi reported that Lakebase has demonstrated early momentum. Within its first eight months on the market, it generated twice the revenue that Databricks’ original data warehouse achieved in its own first eight months. Although he acknowledged that comparing early-stage products has limitations, he described Lakebase as a significantly larger “toddler” in relative terms.

With the latest funding round finalised, Ghodsi stated that Databricks is not actively seeking additional capital or preparing for an IPO at this time. He explained that raising substantial funds was partly a precautionary measure given economic uncertainty.

Referencing the 2022 downturn — when rising interest rates followed a prolonged period of near-zero borrowing costs — Ghodsi emphasised the value of maintaining strong financial reserves. Having significant capital on hand, he said, provides protection and operational flexibility if broader market conditions weaken again, ensuring that the company retains a long runway regardless of volatility.