Strengthening Real-Time AI Platform to Accelerate Corporate Decision-Making
Confluent, a data streaming platform provider, has announced new features on its Confluent Intelligence platform that enable integration of artificial intelligence (AI) agents with business data in real-time. This innovation aims to help companies obtain more accurate data analysis whilst accelerating AI-based decision-making.
The company has introduced Streaming Agents, which use the Agent2Agent (A2A) protocol to connect and coordinate various external AI agents through real-time data streams. With this approach, AI systems across different parts of an organisation can communicate and work collaboratively at an enterprise scale.
Additionally, Confluent has introduced a Multivariate Anomaly Detection feature that analyses multiple metrics simultaneously to detect unusual patterns in data streams. This technology is designed to help IT teams prevent operational disruptions before they impact business systems.
Sean Falconer, Head of AI at Confluent, stated that companies wishing to compete in the digital era require AI systems capable of working collaboratively and responding to data directly. “If you want to compete, your AI cannot only look backwards. You need an AI agent system that works together, continuously learns and shares insights in real-time,” Falconer said in a press statement received on Tuesday, 10 March.
He added that Confluent Intelligence is designed to connect various AI investments and systems owned by companies, enabling AI to automatically respond to data, take action, coordinate systems and escalate to team members when necessary.
Through Streaming Agents, Confluent connects AI agents with real-time data using the Model Context Protocol (MCP) from Anthropic as well as the A2A protocol for inter-agent communication. This system allows AI agents to continuously analyse data from various frameworks and data platforms such as LangChain, BigQuery, Snowflake and Databricks.
Analysis results can then trigger workflows on company platforms such as ServiceNow and Salesforce to directly execute operational actions. This approach enables companies to convert data analysis into action more rapidly.
The A2A support in Streaming Agents also allows companies to build smarter and reusable AI agents, enhance inter-agent communication, and ensure auditability by recording each agent activity in an immutable system log.
Additionally, inter-agent communication is coordinated using Apache Kafka technology, which allows integration of agent output with other systems in the company.
Beyond AI agent integration, Confluent has introduced the Multivariate Anomaly Detection feature, which is part of the built-in machine learning functions on its platform.
Unlike traditional anomaly detection that analyses metrics separately and often relies on historical data, this technology analyses several related metrics simultaneously. This approach can reduce false positives whilst helping IT teams detect problems more quickly.
According to Confluent, this technology can help companies leverage real-time data to increase revenue, reduce risk and lower operational costs across various industries, ranging from retail, financial services, healthcare, manufacturing to telecommunications.
The Sell with Confluent programme aims to simplify partnerships by providing clearer tools and support so partners can sell data streaming solutions more quickly.
In today’s fast-paced world, data streaming provides instantaneous insights into every aspect of business. Data streaming foundations have now become a prerequisite in developing modern AI applications, particularly amid business competition that increasingly depends on data speed and accuracy. With accelerating digital transformation and increasing regulatory pressure on data management, companies now require solutions that can deliver cloud-like speed. Helping customers navigate an increasingly real-time and AI-driven world can only be accomplished with a strong global partner ecosystem.