mlflow-onboarding

To get started with MLflow tracking, the skill detects whether your project needs GenAI tracing or traditional ML experiment tracking, then guides you to the correct quickstart and integration path for your codebase.

To add MLflow observability and tracing to a GenAI app, the skill identifies GenAI code signals like LangChain usage and directs you to specific quickstarts for tracing, evaluation, and prompt versioning integration.

MLflow autologging applies to both workflows, but the skill helps determine when to use autologging versus manual instrumentation by first detecting if your codebase targets traditional ML experiment tracking or GenAI observability.

The best way to set up MLflow experiment tracking for deep learning is detecting traditional ML signals in your codebase, then following targeted onboarding steps for logging parameters, metrics, artifacts, and models.

You can integrate MLflow without manual logging code by using autologging, and the skill provides consent-based integration steps to verify autologging or manual instrumentation fits your specific GenAI or traditional ML setup.

MLflow onboarding determines your tracking needs by inspecting codebase indicators and mlflow experimentKind tags to identify GenAI app signals versus traditional ML or deep learning model training workflows.