AI and ML integration

Use Lakebase as a backend for AI agents and machine learning workloads.

• • Agent state and memory
  • Persist agent state and memory in Lakebase so AI agents built with LangGraph or the OpenAI Agents SDK retain context across sessions.
• • Feature Store and Model Serving
  • Serve ML features at low latency from Lakebase for models deployed with Databricks Model Serving.
• • Lakebase Search
  • Add hybrid vector and keyword search to your project using the lakebase_vector and lakebase_text extensions.

Additional resources

• Lakehouse integration
• Connect to your database
• Query your data