Skip to main content

What's new in Striim Cloud 5.2.0

Looking for the Striim Platform documentation? Click here.

The following features are new in Striim Cloud 5.2.0.

Top new features

  • The AI Agents component in Flow Designer provides prebuilt AI capabilities that can be used as part of streaming data pipelines, with no coding required.

  • The following AI agents are available in the Flow Designer:

    • Sentinel: offers real-time sensitive data detection within streaming applications.

    • Euclid (vector embeddings): enables semantic search, categorization, and enhanced analysis using vector representations.

    • Foreseer (anomaly detection): provides time series forecasting and anomaly detection for real-time monitoring and predictive analytics.

  • Striim supports reading from Db2 for z/OS by Database Reader and Incremental Batch Reader, and change data capture from Db2 for z/OS using an external source. See Db2 for z/OS and Db2 for z/OS CDC..

  • Iceberg Writer supports writing to Apache Iceberg tables in Google Cloud Storage.

  • Striim supports Neon for serverless Postgres and building AI agents. See Database Reader, PostgreSQL CDC reader, and Database Writer.

  • Striim supports Crunchy Postgres. See Database Reader, PostgreSQL CDC reader, and Database Writer.

Web UI

  • Striim AI components are now available in the Flow Designer.

Striim AI features

  • The AI Agents component in Flow Designer provides prebuilt AI capabilities that can be used as part of streaming data pipelines, with no coding required.

  • The following AI agents are available in the Flow Designer:

    • Sentinel: offers real-time sensitive data detection within streaming applications.

    • Euclid (vector embeddings): enables semantic search, categorization, and enhanced analysis using vector representations.

    • Foreseer (anomaly detection): provides time series forecasting and anomaly detection for real-time monitoring and predictive analytics.

  • Optimizing the performance of Striim AI server: Striim AI now uses batch processing and parallel execution instead of handling records individually. You can set performance tuning parameters to optimize this architecture based on your hardware and accuracy needs.

Application development

  • Anomaly detection and time series forecasting: Striim integrates machine learning models, such as time series forecast and anomaly detection, which are implemented as built-in functions. Striim allows you to apply these models to your streaming data for data preprocessing and to analyze your streaming data for automated prediction, and anomaly detection.

Sources and targets

Change data capture