databricks-agent-bricks

Build conversational AI applications with Databricks Agent Bricks by combining Knowledge Assistants, Genie Spaces, and Supervisor Agents. You orchestrate multi-brick deployments using manage_ka and manage_mas, then monitor provisioning until the status reaches ONLINE.

Orchestrate multi-agent workflows in Databricks using Supervisor Agent tiles to coordinate routing and safety. This framework enforces structured configurations and provisioning lifecycles to automate data-driven assistants and policy retrieval end-to-end.

A Genie Space works as a pre-built AI tile for SQL data exploration, enabling users to query data conversationally. It integrates into multi-brick deployments alongside Knowledge Assistants to provide structured data exploration within conversational apps.

Reuse pre-built AI tiles like Knowledge Assistants and Genie Spaces to accelerate Databricks app development. The framework provides best practices and provisioning guidance for data sources, routing, and safety to rapidly assemble workflows.

Provisioning limitations for Knowledge Assistant and Supervisor Agent workflows require monitoring tiles until they reach an ONLINE status. The framework enforces structured configurations for data sources and safety, requiring strict adherence to provisioning lifecycles before deployment.

Structured configurations are required to deploy multi-brick Databricks AI workflows. The framework enforces provisioning lifecycles and best practices for data sources, routing, and safety to ensure reliable orchestration of Knowledge Assistant and Supervisor Agent tiles.