Confluent adds snapshot queries to Apache Flink for agentic AI workloads

Confluent has announced new capabilities for its cloud platform that enable organisations to combine real-time and historical data processing for artificial intelligence agents and analytics.

The company's snapshot queries feature, now available in early access for Confluent Cloud for Apache Flink, allows teams to unify streaming and batch data using a single platform and query language.

Apache Flink is an open-source stream processing framework designed for distributed, high-performance processing of both real-time and batch data.

Flink processes data streams with low latency and high throughput, making it suitable for applications requiring real-time analytics and event processing.

"Agentic AI is moving from hype to enterprise adoption as organizations look to gain a competitive edge and win in today's market," Shaun Clowes, chief product officer at Confluent, said.

"But without high-quality data, even the most advanced systems can't deliver real value," he said.

Banks implementing fraud detection systems exemplify the need for unified data processing, requiring real-time transaction data alongside historical customer patterns to identify suspicious activity.

Previously, organisations had to use separate tools and develop manual workarounds to combine historical and streaming data, resulting in fragmented workflows.

Snapshot queries eliminate this complexity by providing Tableflow integration, enabling teams to explore and analyse data without creating new workloads.

Confluent has also introduced CCN (Confluent Cloud network) routing, now generally available on Amazon Web Services in all regions where Flink is supported.

The feature simplifies private networking by allowing teams to reuse existing CCNs created for Apache Kafka clusters to securely connect data to Flink workloads.

The company has added IP Filtering for publicly accessible Flink pipelines, enabling teams to restrict internet traffic to allowed IP addresses and improve visibility into unauthorised access attempts.

These enhancements complement Confluent's broader platform updates, including a Snowflake source connector and cross-cloud Cluster Linking.

By unifying real-time and historical data processing, organisations can provide AI agents with the comprehensive context needed for accurate decision-making.