mlflow-genai-foundation

What problem does it solve?

This Skill provides foundational patterns for building Databricks GenAI agents, ensuring compatibility with MLflow 3.0 features like model signatures, tracing, and prompt registry.

Core Features & Use Cases

• MLflow Model Signatures: Ensures agents work with AI Playground and other Databricks features by automatically inferring compatible signatures.
• MLflow Tracing: Enables automatic or manual tracing of agent operations for debugging and performance analysis.
• MLflow Prompt Registry: Facilitates versioning and management of prompts used by agents.
• Evaluation Basics: Introduces built-in and custom scorers for evaluating agent performance.
• Use Case: When developing a new GenAI agent for customer support, use this Skill to correctly implement MLflow model signatures, set up tracing for debugging user interactions, and log prompts to the registry for version control.

Quick Start

Implement the MLflow GenAI Foundation patterns for your new Databricks agent.