Cloudera, once a fast-growing Hadoop startup, raised $1 billion and went public in 2018 before being acquired by private equity for $5.3 billion in 2021. Today, the company announced it was acquiring Verta, an AI startup that helps customers manage machine learning models, including large-scale language models used in generative AI.
Cloudera launched a SaaS data lakehouse the year after it was acquired, but to stay relevant in today's market, it needed the power of AI. Cloudera CEO Charles Sansbury certainly recognized that.
“The future of data management is AI and the two go hand in hand. Cloudera is acquiring Verta's operational AI platform to strengthen our team and accelerate our operational AI capabilities,” he said in a statement.
As enterprises moved to large language models, Verta evolved from a task-based model management platform that acted as a control center for models to the platform it is today, purpose-built for managing large language models.
At a time when quality AI talent is hard to come by, the acquisition gives Cloudera top-tier talent to help operationalize and scale its AI tools, including MIT CSAIL-trained co-founders CEO Manasi Vartak and CTO Conrado Miranda, formerly the machine learning leader at Twitter.
Verta was founded in 2018 and has raised about $16 million, according to Pitchbook, including a $10 million Series A in 2020. Vartak created the open-source project ModelDB Database while in grad school as a way to track versions of machine models; she later expanded that idea into Verta.
Cloudera began life as a Hadoop startup in 2008, at a time when companies were just starting to think about how to process huge amounts of data. Hadoop, an open-source project developed at Yahoo in 2005, was once the cutting edge way to process data. The problem was that by the time the company went public, there were simpler, more cost-effective ways to process that data, and Hadoop had lost steam.
At the same time, companies were moving many of their data workloads to the cloud, whether it was the big three cloud vendors — Amazon, Microsoft, and Google — or startups like Snowflake and Databricks. Despite the name, Cloudera's solution was actually on-premise for most of its existence.
The move to build out a SaaS data lakehouse in 2021 was also an attempt to compete with cloud-native competitors, and since then, both Databricks and Snowflake have added AI capabilities both organically and through acquisitions.
Today's move is really about keeping up with the Joneses.