agent-bricks

What problem does it solve?

This Skill helps teams accelerate the development of conversational AI applications by bundling three pre-built Databricks components—Knowledge Assistants (KA), Genie Spaces, and Multi-Agent Supervisors (MAS)—into reusable agent bricks.

Core Features & Use Cases

• KA integration: Build document-based Q&A experiences using Knowledge Assistants backed by Unity Catalog data volumes.
• Genie Space guidance: Convert natural language queries into SQL against Unity Catalog tables for exploratory analysis.
• MAS orchestration: Route questions to specialized agents and endpoints, enabling scalable, multi-agent workflows.
• End-to-end workflow: Provision KA, Genie Space, and MAS with tool-based operations and provisioning status monitoring.

Quick Start

Begin by creating a KA with create_or_update_ka, then create a Genie Space with create_or_update_genie, and finally assemble a MAS with create_or_update_mas. Monitor provisioning status and add sample questions once endpoints are ONLINE.