Confluent Adds Query and Real-Time AI Tools to Its Streamhouse Platform

Query turns Kafka data into live business state

Confluent says Query joins Stream, Connect, Process and Govern as the fifth Data Streaming Platform pillar. Its central feature, Lightning Tables, is a fully managed, read-only serving layer that makes Kafka data queryable with low latency and high concurrency.

According to Confluent, Lightning Tables can provide agents, applications and analytics engines with fresh business context without requiring a separate serving database. The company says Kafka streams can be turned into a real-time state of the business, allowing agents and applications to query what is true now in milliseconds, with no additional database or ETL.

The update also expands Tableflow. It now supports Google Cloud and BigQuery, with automated BigLake Metastore synchronization. Confluent says this allows teams to query live Kafka data in minutes without pipeline code.

Managed inference brings AI closer to streaming data

Managed Model Inference is now generally available. The zero-configuration capability lets teams call supported AI and machine learning models from Flink SQL without managing credentials, endpoints or serving infrastructure.

Confluent says managed inference is generally available for large language models and embedding models hosted in Amazon Bedrock, Vertex AI and Azure OpenAI. It abstracts serving layers, credentials and hardware management, which the company says can help teams get started faster with AI on streaming data.

The Real-Time Context Engine is also generally available. It serves fresh context to AI agents through MCP as an access pattern for Lightning Tables. The release includes bring your own key support, OAuth through Confluent Cloud sign-in or enterprise single sign-on, metrics API visibility into storage bytes, bytes scanned and CCU minutes, and support for backward-compatible schema changes.

AI assistants can explore, query and operate Confluent environments

Confluent’s managed MCP servers are generally available for use with AI assistants. They allow teams to explore environments and clusters, inspect topics and schemas, debug, and query. New Flink SQL tools can create, list, read and delete statements, debug statements, retrieve metrics and look up table structure.

OAuth provides one Confluent Cloud sign-in, without an API key, across global and regional MCP servers and the Real-Time Context Engine. Confluent says these AI tools are free for Confluent Cloud users working with assistants such as Claude, Cursor and Copilot.

Confluent is also offering Confluent Skills as generally available coding-assistant plugins. These guided workflows help developers design, build and debug Kafka Streams applications. The Confluent Tuning Advisor inspects Kafka configurations and recommends tuning changes without applying them.

For users working directly in Confluent Cloud, Confluent Assistant can diagnose fully managed connectors that have entered a FAILED state with a runtime error. With explicit user consent, it can apply a recommended fix and verify that the connector returns to a RUNNING state. The company says the feature is available for early access.

Together, the changes give Confluent users more ways to work with live streaming data, from managed inference and agent context to natural-language operations. The announcement also includes Schema Registry updates, multi-tenancy for Confluent Private Cloud, Confluent Gateway enhancements and automatic failover for Kafka in Confluent Cloud, extending the platform’s focus on shared, governed and continuously updated business state.