Databricks Inc. today put some meat on the bones of its announcement three weeks ago that it had raised new funding that valued the company at more than $100 billion.

The company said it is closing a Series K venture capital round of $1 billion, bringing its total funding to over $20 billion. Databricks said it would use the capital to expand the market for its Agent Bricks and Lakebase initiatives and expand globally.

Agent Bricks is a unified workspace for building and optimizing artificial intelligence agents. Lakebase is a serverless database designed to handle both transactional and analytical workloads. Both were announced in June.

The company also released some details about its sales and growth. Databricks crossed the $4 billion revenue run-rate threshold in the second quarter, up 50% year-over-year. Sales of its AI products alone surpassed the $1 billion revenue run-rate.

Databricks said more than 650 customers spend more than $1 million annually on its products, and it has an impressive net revenue retention rate of over 140%. NRR is a metric that shows how much existing customers change their spending annually. The company also said it has maintained positive free cash flow over the past year, an important indicator of financial health.

Needs the money?

Although the size of the company’s war chest may raise questions about why it needs more money, Databricks’ shift to serverless delivery earlier this year has probably taken a hit on expenses, said Rob Strechay, an analyst at theCUBE Research.

“They’re going all in on serverless infrastructure, but that can be significantly more expensive depending on how you utilize it,” he said. Databricks also launched a free version in June, which has added to the cost of goods sold without a corresponding revenue benefit.

“The combination of serverless and the free tier could be creating a fog across their financial modeling,” Strechay said. At its current market, raising $1 billion has little offsetting cost. “Why not take the money when it’s cheap?”

Enterprise interest in AI is a clear tailwind. Just last month, the company said it was projecting an annual run rate of $3.7 billion. That has now been updated to $4 billion less than a month later.

Databricks executives have hinted about plans for an initial public offering since 2021 but have lately shown little urgency about making that move. The $100 billion valuation puts the company well ahead of rival Snowflake Inc.’s $76.4 billion market capitalization; however, the comparison is a bit of an apples-and-oranges exercise, said David Vellante, chief analyst at theCUBE Research.

Plenty of opportunity

“Remember that Databricks is a private company and subject to far less investor scrutiny,” he said. “Investors are betting on Databricks’ future, whereas Snowflake faces a more liquid market of buyers and sellers.”

Although much has been made of the two companies’ rivalry, “We don’t see Databricks versus Snowflake as a zero-sum game,” Vellante said. “Rather, we see the market as sizable, and both companies can thrive.”

Strechay said the buffeting Snowflake has taken in the market since its IPO four years ago may have highlighted the virtues of staying private for Databricks executives. “I think they like getting their ducks in a row versus having to report every quarter what’s going on underneath the hood,” he said.

A Databricks spokeswoman said the company’s official stance is “We’ll go public when the timing makes sense. We are already acting like a public company in many ways and will be ready when the time comes.”

The latest funding round was co-led by Andreessen Horowitz LLC, Insight Partners Inc., MGX Fund Management Ltd., Thrive Capital Management LLC, and WCM Investment Management LLC.

Photo: Robert Hof/SiliconANGLE