It was only eight months ago that machine learning and Apache Spark specialist Databricks raised $250 million in a Series E funding round that valued it at $2.75 billion, drawing in capital from Microsoft, as well as long-standing backers Anderson Horowitz.
Building up the firm’s UK presence, CEO Ali Ghodsi told Computer Business Review at the time that “the emphasis is on growth: we have a unique sales strategy focused on industry verticals and this investment will allow us to broaden the number of verticals we’re talking to.”
The company, which names FIS, HP and Shell as customers, took the opportunity to ramp up its EMEA presence, opening an office on London’s Baker Street in the former House of Fraser building.
Now, less than nine months later, its investors like what they see so much they’re injecting another $400 million. The Series F Databricks funding round values the firm at a huge $6.2 billion valuation.
While it claims annual recurring revenue (ARR) of well over 2.5x over the past year, it’s a bullish valuation. So, what’s the plan with all the money?
Databricks says it will:
- Invest €100 million in its Development Center in Amsterdam; an engineering hub that has grown three-fold in two years.
- Build dedicated engineering teams for Delta Lake, MLflow, and Koalas.
- “Fuel expansion in EMEA; APAC and LatAm.
The company offers of four open source products: Delta Lake, a data lake product; MLflow, a tool to deploy machine learning; Koalas, which creates a single machine framework for Apache Spark and pandas; and, finally, Spark itself, the analytics engine that Databricks’ founders created at Berkeley.
(Ghodsi told Computer Computer Business Review last year that when he and his colleagues first built Spark, they couldn’t even give it away. He told us: “We put it on a few web pages and said: ‘Hey download this!’ No one downloaded it. So we went to existing Hadoop vendors and many tech companies and said: ‘Hey can you adopt this technology?’ They all told us ‘no: this is academic mumbo-jumbo. We need enterprise software.’ We literally wanted to give it away and no one would take it.”
With a fresh $400 million in the company’s bank accounts, eBay and Netflix among those now deploying Spark at massive scale to process petabytes of user data and investors including Andreessen Horowitz, Blackrock, Geodesic, Microsoft and more keen to keep injecting funds, Ghodsi and team have clearly had the last laugh.
“Databricks’ net revenue retention is astounding,” said David George, general partner at Andreessen Horowitz. “Why? Because customers love the product. Beyond that, they’ve followed up their breakthrough Spark technology with an amazing series of open source innovations.”
Our entirely speculative prediction? Microsoft will make a move for the company within 18 months.