IBM has agreed to buy DataStax, a provider of managed and cloud-hosted Apache Cassandra database and Pulsar data streaming services.
Principal technical strategist at DataStax Patrick McFadin said the deal “signals a strategic bet on the future of Cassandra” (an open-source NoSQL database deployed by the likes of Apple and Netflix.)
The deal is something of a triumph for CEO Chet Kapoor, who joined five years ago at a low point for DataStax. This will be the second company that he has sold to IBM; Kapoor was CEO of Gluecode when IBM bought it in 2005. As CEO he also successfully sold Apigee to Google in 2016.
DataStax’s blue chip customers include Fedex, Home Depot, and T-Mobile. The terms of the deal were not initially disclosed. DataStax employees have been promised an information pack “later” today (February 25). The Stack will update this article if we hear more details on deal size.
IBM’s proposed DataStax acquisition follows its $34 billion buyout of Red Hat in 2019 and its $6.4 billion bid* for Terraform creator HashiCorp – which British competition regulators approved today.
“The project needs a lot of friends”
The proposed deal will leave many in the open-source community somewhat anxious: DataStax has had a complex relationship with the OSS Cassandra community but in recent years rebuilt partnerships and been a sustained, active contributor upstream, even as it also builds premium “enterprise” features separately for its “Astra” set of managed services.
Ritika Gunnar, General Manager, Data and AI, IBM, played up Big Blue’s commitment to open-source in a blog on the news, writing: “At IBM, we believe in the power of open source and the innovation driven by community collaboration… We share this belief with DataStax, who has continued to expand their stewardship of open source including Apache Cassandra, Langflow, OpenSearch, and more… we look forward to supporting and innovating alongside these communities, helping shape the future of enterprise data management and AI-driven solutions.”
Posting on LinkedIn, McFadin wrote: “In 2011, the realistic question was whether Cassandra would survive the expansion and contraction of NoSQL. In 2025, the new question is whether Cassandra can be the default database for a generation. To get there, the project needs a lot of friends, and it has some great ones. Netflix, Apple, Bloomberg, Uber.
“It’s a long list. Today, IBM joined the chorus. The Cassandra project momentum has been picking up and just got some big blue jet boots.”
See: "DataStax Astra took Cassandra serverless. That's a game-changer"
DataStax CEO Chet Kapoor wrote: “DataStax’s NoSQL products—our cloud service, Astra DB, and self-managed DataStax Enterprise—have been the standard for scale-out data delivered on top of Cassandra.
“DataStax and IBM have collaborated in the market since 2020, serving customers such as T-Mobile, Audi, The Home Depot, and Intuit.
“Over the last year, DataStax introduced HCD and Mission Control, further bringing Cassandra into the cloud-native era, deployed on top of IBM OpenShift. We’re excited to continue making Cassandra the standard for scaling out cloud-native data for our existing and future customers.”
The deal follows DataStax’s April 2024 acquisition of Langflow, which provides a GUI for building retrieval augmented generation and other AI applications with LangChain. IBM said it sees Langflow as helping it add middleware development capabilities to its watsonx.ai platform.
It comes amid a busy week in the database space, with MongoDB yesterday unveiling plans to buy AI model/RAG specialist Voyage AI for $220 million in a cash and shares deal.
Reaction and more details to follow. Views? Pop us a line.
*That latter deal was dubbed “Red Hash” by some; a moniker that arguably makes this one “Red Cass”, or perhaps “Red Stax"; take your pick.
