Evaluating Supervisor Agents with MLflow on Databricks

In this video we expand on the Multi-Agent Supervisor option we explored in Databricks, by showcasing how you can evaluate this option utilizing MLFlow constructs. Video Resources - Previous Video (Create Supervisor Agent): https://www.youtube.com/watch?v=BRnGydDiCIo&t=544s - Notebook Code: https://github.com/RamVegiraju/databricks-samples/tree/master/foundation-models/Agents/AgentBricks/ManagedSupervisor - MLFlow Judges and Scorers Docs: https://mlflow.org/docs/latest/genai/eval-monitor/scorers/ Timestamps 0:00 Introduction 1:50 Evaluating GenAI Systems 5:10 MLFlow GenAI Scorers & Judges 10:50 Evaluating Supervisor Agent 13:10 What is a Trace 14:40 Hands-On #databricks #agentbricks #agents #genie #mlengineering #supervisoragent #agentevaluation #mlflow