IBM to buy Confluent to extend its data and automation portfolio

IBM has agreed to acquire cloud-native enterprise data streaming platform Confluent in a move designed to expand its portfolio of tools for building AI applications

The company said Monday in a release that it sees Confluent as a natural fit for its hybrid cloud and AI strategy, adding that the acquisition is expected to “drive substantial product synergies” across its portfolio.

Confluent connects data sources and cleans up data. It built its service on Apache Kafka, an open-source distributed event streaming platform, sparing its customers the hassle of buying and managing their own server clusters in return for a monthly fee per cluster, plus additional fees for data stored and data moved in or out. 

IBM expects the deal, which it valued at $11 billion, to close by the middle of next year.

Confluent CEO and co-founder Jay Kreps stated in an email sent internally to staff about the acquisition, “IBM sees the same future we do: one in which enterprises run on continuous, event-driven intelligence, with data moving freely and reliably across every part of the business.”

It’s a good move for IBM, noted Scott Bickley, an advisory fellow at Info-Tech Research Group. “[Confluent] fills a critical gap within the watsonx platform, IBM’s next-gen AI platform, by providing the ability to monitor real-time data,” he said, and is based on the industry standard for managing and processing real-time data streams. 

He added, “IBM already has the pieces of the puzzle required to build and train AI models; Confluent provides the connective tissue to saturate those models with continuous live data from across an organization’s entire operation, regardless of the source. This capability should pave the road ahead for more complex AI agents and applications that will be able to react to data in real time.”

He also pointed out that the company is playing the long game with this acquisition, which is its largest in recent history. “IBM effectively positions itself proactively to compete against the AI-native big data companies like Snowflake and Databricks, who are all racing towards the same ‘holy grail’ of realizing AI agents that can consume, process, and react to real-time data within the context of their clients’ trained models and operating parameters,” he said, adding that IBM is betting that a full-stack vertical AI platform, watsonx, will be more appealing to enterprise buyers than a composable solution comprised of various independent components.

The move, he noted, also complements previous acquisitions such as the $34.5 billion acquisition of Red Hat and the more recent $6.4 billion acquisition of Hashicorp, all of which are built upon dominant open source standards including Linux, Terraform/Vault, and Kafka. This allows IBM to offer a stand-alone vertical, hybrid cloud strategy with full-stack AI capabilities apart from the ERP vendor space and the point solutions currently available.

In addition, he said, the timing was right; Confluent has been experiencing a slowing of revenue growth and was reportedly shopping itself already.

“At the end of the day, this deal works for both parties. IBM is now playing a high-stakes game and has placed its bet that having the best AI models is not enough; it is the control of the data flow that will matter,” he said.