Confluent, an IBM company and the data streaming pioneer, has announced new capabilities in Confluent Intelligence and Confluent Cloud. The updates aim to streamline how real-time artificial intelligence applications are built and secured.

The company said these updates remove key security and complexity barriers that often prevent organizations from moving AI workloads into real-world production environments. Confluent noted that its platform unifies the AI lifecycle with tools developers already use.

In addition, it integrates Apache Flink pipelines with dbt, also known as data build tool. It also introduces a fully managed Model Context Protocol server and Agent Skills. These features allow AI systems to manage streaming operations more effectively.

Moreover, Confluent has added automated personally identifiable information redaction. It also enables private connectivity to external models through Azure Private Link. As a result, enterprise-grade governance is embedded directly into data streams.

Sean Falconer, Head of AI at Confluent, said most AI projects fail before reaching customers because the data layer breaks down. He added that teams often have models and intent, but security risks and fragmented data prevent deployment. Therefore, the company aims to make the streaming layer a foundation for secure, production-ready AI.

According to McKinsey, eight in ten companies cite data limitations as a barrier to scaling agentic AI. This is often due to security teams restricting data access because of exposure risks. At the same time, developers lose significant time switching tools to manage and inspect streaming data. Consequently, AI development becomes slower and more manual.

Confluent Cloud and Confluent Intelligence together form a data streaming foundation for production-ready AI. They continuously process historical and real-time data. Then, they deliver trusted context into AI applications.

Furthermore, natural language operations allow developers to use Confluent MCP as a control plane. This enables AI to build, manage, and debug streaming workflows using natural language. Agent Skills also encode best practices and workflows. Together, they ensure consistent execution aligned with organizational standards. These features are generally available for Confluent Cloud.

In addition, automated data privacy is supported through a built-in machine learning function for PII detection and redaction in Flink SQL. This works without custom code or external services. It also avoids moving data to a warehouse first. The feature is in early access for Confluent Intelligence.

Secure connectivity is also enhanced through Azure Private Link support. This keeps AI workloads off the public internet. It provides private connectivity to external models and data sources. As a result, Flink jobs can connect securely to Azure OpenAI, Azure SQL, and Cosmos DB over Microsoft’s private backbone. This capability is generally available on Confluent Cloud.

At the same time, unified engineering workflows are enabled through a free open source dbt adapter. It brings Flink SQL on Confluent Cloud into dbt, a widely used framework for building data pipelines. This allows teams to define, test, and deploy streaming pipelines using existing dbt structures and commands. It reduces barriers to Flink adoption and extends real-time capabilities into existing workflows. This is generally available on Confluent Cloud.

Additionally, Confluent supports TimesFM models for anomaly detection. It also supports Anthropic and Fireworks AI models. Developers can use these directly in Flink stream processing workflows to build real-time AI applications.

Highlights include the general availability of the Real-Time Context Engine. This engine continuously delivers fresh and governed context for AI applications. New fully managed connectors in Confluent Cloud also simplify data integration.

These updates extend recent announcements at IBM Think. They integrate Confluent Cloud further into IBM solutions. With Confluent, watsonx.data provides an AI-ready data foundation and real-time context layer for hybrid environments.

Finally, Confluent said these capabilities strengthen its role in secure and scalable AI infrastructure. The company emphasized that Confluent continues to power real-time intelligence across enterprises.

Related post

View all