Confluent, Inc., the data streaming pioneer, announced the launch of Streaming Agents, a new capability in Confluent Cloud for Apache Flink®. The solution enables enterprises to build and scale agentic AI with real-time data access, empowering AI agents to monitor, reason, and act effectively. By unifying data streaming and AI workflows, Streaming Agents accelerates the adoption of enterprise-grade agentic AI.

According to Confluent, Streaming Agents simplify integration with secure connections across business systems, large language models, and tools. As a result, organizations build a stronger foundation for deploying AI agents. This approach reduces delays in implementation. It also unlocks new opportunities for digital transformation.

“Agentic AI is on every organization’s roadmap, yet many remain stuck in prototype purgatory,” said Shaun Clowes, Chief Product Officer at Confluent. “Streaming Agents delivers the missing context AI agents need by unifying real-time data with intelligent workflows.”

Industry Adoption of Agentic AI Still Faces Barriers

IDC research highlights the challenges of bringing agentic AI into production. On average, organizations tested 23 generative AI proofs of concept between 2023 and 2024. However, only three reached production, and just 62% met expectations. Complex workflows and limited real-time integration remain major obstacles.

Stewart Bond, Vice President of Data Intelligence and Integration Software at IDC, emphasized that most enterprises lack data architectures that support autonomous AI decision-making. “Enterprises need agentic AI solutions with secure integration and real-time context to deliver intelligent action,” Bond said.

Streaming Agents Unlock Real-Time AI Capabilities

Streaming Agents embeds agentic AI into data pipelines, allowing enterprises to deploy event-driven agents with Apache Kafka® and Apache Flink®. The capability ensures that AI agents continuously access fresh, contextual data and collaborate with other systems as business conditions evolve.

For example, Streaming Agents can track competitor pricing across e-commerce sites and automatically adjust a retailer’s prices in real time. This enables enterprises to deliver competitive offers instantly while reducing manual intervention.

Key Features of Streaming Agents Include:

  • Tool Calling for Context-Aware Automation: Through Model Context Protocol (MCP), agents dynamically select tools such as APIs, SaaS applications, or databases to take relevant actions.
  • Secure Integrations: Enterprises can securely connect to LLMs, vector databases, and MCP while safeguarding sensitive credentials. Centralized connections improve scalability and management.
  • External Tables and Search for Accuracy: By enriching data with relational databases or REST APIs, agents deliver more accurate reasoning, enhanced vector search, and RAG applications.
  • Replayability for Safer Iteration: Developers can test and refine AI agents using real data without live side effects, enabling A/B testing, dark launches, and faster innovation cycles.

With this release, Confluent reinforces its leadership in real-time AI infrastructure, enabling enterprises to scale agentic AI securely and efficiently.

Stay informed on IT and data-driven innovations – visit IT Tech News today.

News Source: Businesswire.com