Confluent Launches Streamhouse Features for AI Agents
The new capabilities extend Confluent's streaming data platform so AI agents can query live, governed data without separate pipelines.
Updated on Nov. 4, 2026 in Artificial Intelligence
Confluent launched new features on Wednesday that turn its streaming data platform into a streamhouse for AI agents. The update includes Lightning Tables, now generally available on AWS, which gives agents live, governed tables.
Why it matters
Customers previously could capture and process streaming data but had to copy it to another platform to query it, adding delay and complexity. As companies build software that runs business operations rather than assisting people, real-time governed data access becomes central to agent performance.
Confluent's Query suite now includes Lightning Tables, joining a real-time context engine, Snapshot Queries, and Tableflow within a five-capability architecture of Stream, Connect, Process, Query, and Govern. Confluent has not published performance benchmarks for Lightning Tables at scale, and the feature is currently generally available only on AWS.
The players
Confluent
Confluent is a data streaming platform company whose technology underpins real-time data pipelines used by enterprises, and it was acquired by IBM in December 2025.
IBM
IBM is a multinational technology company that acquired Confluent in December 2025, bringing its streaming data platform into its enterprise software portfolio.
The details
Lightning Tables is a feature within Confluent's Query suite that provides agents with live, governed tables, eliminating the need to build pipelines and move data to a separate platform for querying. Confluent also introduced hybrid connectors in early access.
Timeline
December 2025: IBM acquired Confluent.
September: A working group including Aiven, Confluent, and Redpanda first introduced the streamhouse concept.
Wednesday: Confluent launched the new streamhouse features.
The Tech Race
The streamhouse concept merges streaming data platforms with data lakehouse architectures, a combination first proposed in September by a working group including Aiven, Confluent, and Redpanda. Confluent's move positions it against rivals building real-time data infrastructure for AI agents.
Developers building AI agents can now query live, governed data without copying it to a separate platform, which may reduce pipeline work. Adoption depends on availability beyond AWS, which Confluent has said it will expand.
The takeaway
Real-time, governed data access is becoming a prerequisite for AI agents that run business processes. Teams evaluating agent infrastructure should assess whether their streaming and query layers can serve live data without extra copies.
Further reading
Explore more on Artificial Intelligence to see how real-time data is shaping agent development.
