Databricks recently completed one of the most significant funding rounds, securing an astonishing $10 billion in new funds. Naturally, technology financiers swiftly inquired about the implications for the company’s eagerly awaited IPO. At a gathering in San Francisco on Tuesday evening, Databricks CEO Ali Ghodsi elaborated on the reasons for postponing a public offering until at least 2025.

“This year marked an election period. We wanted to establish some stability — there are concerns about interest rates, inflation… So we thought, it’s unwise to IPO this year, so we’ll certainly hold off,” stated Ghodsi in a chat with Dan Primack at the Axios AI Summit. “The earliest theoretical option for an IPO could be next year, and then there are lock-up durations, so it would simply be too extended a timeframe for employees to attain liquidity.”

Databricks is utilizing this “Series J” to allow early staff to exit and maintain growth. Although 2024 holds many uncertainties, the IPOs of ServiceTitan, Reddit, and similar entities have been largely triumphant.

But why gamble when raising funds as Databricks does?

Ghodsi remarked that this latest funding round could have been almost twice the amount recently secured. Investors were eager to participate, but their enthusiasm caused Databricks to elevate its share price. Beginning with a goal to procure $3 billion to $4 billion in this round, according to Ghodsi, media coverage of their fundraising ambitions sparked interest exponentially.

“I was presented with an Excel sheet listing everyone eager to invest. It was $19 billion of interest, and my typical reaction was one of surprise,” Ghodsi explained. “And we hadn’t even engaged with everyone. I was like, ‘Wow, that’s a significant figure.’ So, we actually increased the share price.”

Despite the impressive capital raise, Ghodsi hasn’t excluded the possibility of a Databricks IPO in 2025. Nevertheless, he mentioned it might extend to 2026. He indicated it’s considerably less crucial to go public compared to a decade or two ago, as illustrated by this unprecedented round, but it’s something the company aspires to accomplish. That said, Ghodsi isn’t in a rush to finalize an IPO before the “AI bubble,” as he termed it, bursts.

“I mean, it’s the height of the AI bubble. It doesn’t require brilliance to realize that a company with merely five individuals, lacking a product, innovation, or intellectual property, consisting only of recent graduates, isn’t typically valued at hundreds of millions, even billions,” Ghodsi observed. “You witness startups reaching billion-dollar valuations while having nothing substantial — that’s a bubble.”

Ghodsi did not specify which startups he referred to, yet there has undeniably been a plethora of AI unicorns observed this year. None of this seems to disturb Ghodsi, as he believes his organization and its valuation remain enduring. He thinks his firm has already emerged victorious in its initial significant competition with another data analytics company, Snowflake.

“We had a project named ‘SnowMelt,’” Ghodsi acknowledged while confirming strategies within Databricks to entice business from Snowflake. “Our target was Snowflake, and we vilified them, but that is in the past.”

This attempt to malign Snowflake was costly, reportedly leading Databricks to set aside $2 billion to acquire a minor startup, Tabular. Snowflake supposedly also wanted to purchase Tabular, although the firm was only generating $1 million in yearly recurring revenues then.

Currently, Databricks is pursuing larger rivals with offerings that compete with major enterprises like Salesforce and Microsoft. Ghodsi conveys that data and AI will increasingly influence individuals’ lives annually and believes the company is strategically placed to occupy that gap.