On Monday, Databricks announced that it had reached the level of spending of 5.4 billion dollars, growing by 65% year over year, of which more than 1.4 billion dollars came from its AI products.
Founder and CEO Ali Ghodsi wanted to share these growth numbers because there is so much talk about how AI will kill the SaaS business, he told TechCrunch.
“Everybody’s like, ‘Oh, it’s SaaS. What’s going to happen to all these companies? What’s AI going to do to all these companies? For us, it’s increasing usage,” he said.
To be sure, he also wants to divert the SaaS label from Databricks, which is valued as it is in private markets as an AI company. Databricks on Monday also officially closed its largest, previously announced raising of 5 billion dollars at a value of 134 billion dollars, and held a loan facility of 2 billion dollars.
But this company is struggling in the world. Databricks is still well known as a cloud data store provider. A data warehouse is where businesses store large amounts of data to be analyzed for business insights.
Ghodsi named, in particular, one AI product that drives the use of the data warehouse: LLM’s user interface called Genie.
Genie is an example of how a SaaS business can replace its interactions with natural language. For example, you use it to ask why warehouse usage and revenue spike on certain days.
A few years ago, such a request would have required some query language, or perhaps a special report would have been prepared. Today, any product with an LLM interface can be used by anyone, notes Ghodsi. Genie is one of the reasons for the growth in the use of mathematics, he said.
The threat of AI in SaaS isn’t, as one AI VC tweeted jokingly, that businesses will rip up their SaaS “systems of record” to replace them with homegrown versions with a vibe. Systems of record store important business data, whether in sales, customer support, or finance.
“Why would you move your recording plan? You know, it’s hard to move,” said Ghodsi.
Model makers don’t provide details on how to store that data and have systems of record anyway. Instead, they hope to replace the user interface with natural language that humans will use, or APIs or other agent plug-ins.
So the threat to SaaS businesses, Ghodsi says, is that people who no longer use their jobs become masters of a particular product: Salesforce experts, or ServiceNow, or SAP. If the interface is just language, the products become invisible, like pipes.
“Millions of people around the world were trained in those areas of use. And that was a big drain for those businesses,” warned Ghodsi.
SaaS companies that adopt the new LLM interface can grow, as Databricks did. But it also opens up opportunities for native AI competitors to offer more efficient alternatives to AI and agents.
That’s why Databricks created its Lakebase database designed for agents. You see the pull quickly. “In the eight months we’ve had it on the market, it’s made twice as much money as our savings when it was eight months old. Okay, obviously, that’s like comparing toddlers,” said Ghodsi. “But this is a small child twice as big.”
Meanwhile, with Databricks closing in on its biggest funding round, Ghodsi tells us the company isn’t immediately working on another raise, or preparing for an IPO.
“Now is not a good time to go public,” said Ghodsi. “I wanted to make a really good profit” if the markets went “south” again as they risked the 2022 ZIRP. A thick bank account “protects us, gives us many years of runway,” he added.
