Instrument your app: Tracing approaches

Learn about the different approaches you can take to add traces your generative AI application. MLflow Tracing provides end-to-end instrumentation to give you a complete picture of how your app behaves.

MLflow has three approaches to tracing.

• Automatic - Add one line mlflow.<library>.autolog() to automatically capture app logic for 20+ supported libraries.
• Manual - Designed for custom logic and coplex workflows, control what gets traced using using high-level APIs or low-level APIs.
• Combined - Mix both approaches for complete coverage.

Which approach should I use?

Start with automatic tracing. It's the fastest way to get traces working. Add manual tracing later if you need more control.

For example, you could use the auto-tracing for OpenAI's SDK and manual tracing to combine multiple LLM calls into a single trace that represents your application's end to end logic.

Determine the best tracing approach for your use case based on how you are writing your application's code:

Using one GenAI library (LangGraph, CrewAI, OpenAI Agents, Bedrock Agents, and others)

• Use automatic tracing for your selected library: mlflow.<library>.autolog()

Using LLM SDKs directly (e.g., OpenAI SDK, Anthropic SDK, Bedrock SDK, etc)

• Use automatic tracing for the API library
• Add manual tracing decorators to combine multiple LLM calls into a single trace

Using multiple GenAI libraries or SDKs (e.g., LangGraph AND OpenAI SDK, etc)

• Enable automatic tracing for each framework / SDK
• Add manual tracing decorators to combine calls to multiple frameworks or SDKs into a single trace

All other approaches or you have a need more control

• Use manual tracing
  • Start with the high-level APIs (@mlflow.trace decorator and fluent context managers) which provide a balance of control and ease of use
  • Use the low-level APIs only if the high-level APIs don't give you enough control

Next steps

• Automatic Tracing - Explore all supported libraries and frameworks for one-line integration
• Manual Tracing - Learn advanced techniques for custom instrumentation
• Debug & observe your app - Use traces for debugging and monitoring

Reference guides

Explore detailed documentation for concepts and features mentioned in this guide.

• Tracing concepts - Understand the fundamentals of MLflow Tracing
• Tracing data model - Learn about traces, spans, and attributes
• Query traces - Programmatically access and analyze your traces