IBM snaps up Confluent in $11B all-cash deal

IBM has entered into a deal to buy Kafka data-streaming company Confluent, the companies announced Monday morning.

The price is $31 per share, a 25.5 percent premium over the $23.14 share price, with an $11 billion transaction value. Confluent CEO Jay Kreps told employees: “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.”

IBM chairman, president, and CEO Arvind Krishna said: “IBM and Confluent together will enable enterprises to deploy generative and agentic AI better and faster by providing trusted communication and data flow between environments, applications and APIs. Data is spread across public and private clouds, datacenters and countless technology providers. With the acquisition of Confluent, IBM will provide the smart data platform for enterprise IT, purpose-built for AI.”

The companies note that in the last four years, Confluent’s total addressable market (TAM) has doubled from $50 billion to $100 billion in 2025. Its real-time data and event streaming capabilities, combined with IBM’s AI infrastructure software and automation offerings, will better position the companies to capture this opportunity.

Data infrastructure is important. By buying Confluent, IBM will be getting a substantial real-time data ingestion and streaming engine and platform that it can incorporate in its AI data pipeline offerings and feed its watsonx agents. IBM bought DataStax in May for an estimated $3 billion, also acquiring NoSQL and vector databases based on Apache Cassandra and the Langflow open source tool and community for low-code AI application development. Salesforce bought data infrastructure company Informatica for around $8 billion in May, and Databricks announced a structured streaming product in October, which could limit Confluent’s market.

Confluent was founded in 2014 by three ex-LinkedIn engineers who had developed the Apache Kafka real-time data streaming software. Kreps, CTO Jun Rao, and chief architect Neha Narkhede wanted to commercialize Kafka, and launched a self-managed Confluent Platform in 2015. They launched the fully managed Confluent Cloud SaaS offering in 2020, and went public in June 2021, raising $828 million at a $4.5 billion valuation. The stock price opened at $46.15/share, with shares peaking at $93.60 in November that year, but then dropping to the low $20 area in mid-2022. They have been range bound between $20 and $38 since then.

Confluent is capitalized at $8.09 billion, with shares trading at $23.14 on December 5 down from the year’s high point of $37.65 in February, helped by AI excitement that has since waned. The share price dropped to $15.91 in August.

It reported its Q3 calendar 2025 results in October. These showed it having $286.3 million in subscription revenues and a $1.2 billion run rate, with a $66.5 million GAAP loss. The company has been steadily growing revenues in the double-digit area and gradually reducing its losses as a proportion of revenue since the start of 2023, but profitability is a distant prospect. Here’s a revenue and loss history chart:

The chart of its stock price over the same period shows investors didn’t think that much of its steady growth, even with quarterly revenues typically rising double digits year-over-year and the controlled GAAP losses.

Reuters reported in October that Confluent was looking to be acquired. It became vulnerable to takeover approaches when its stock price dived in July after it reported losing business from a large customer. That customer moved from using Confluent’s cloud offering to having its own self-hosted, on-premises capability.

Although AI has boosted the data infrastructure business, most of the excitement is in the data lake/lakehouse area, exemplified by Databricks and Snowflake. Data infrastructure is needed to underpin this but is less glamorous.

Confluent CEO Jay Kreps talked about GenAI in the October earnings call, saying: ”There’s a few opportunities around AI for Confluent. One is around making the agents real time. One is about the provisioning of real-time data sets. Both of those are actually substantial, and you can do them both together or you can do them separately… We’ll have some announcements in this space that I think will fill out the picture a bit more.”

It announced Confluent Intelligence in November. This combined data streaming, processing, and AI reasoning, enabling developers to build agentic AI systems that act on real-time data. It streams structured, reliable context to AI agents or applications via Kafka, Flink, or MCP. Developers can build and run event-driven AI agents directly on Flink for intelligent, context-aware automation. The agents are natively integrated with Anthropic’s Claude models.

The deal is expected to close by mid-2026, subject to regulatory approvals.