Dynatrace AI Observability & MCP Workshop#
Master AI/LLM monitoring with Dynatrace and the Model Context Protocol (MCP) in this hands-on workshop.
Workshop Overview#
| Duration | 1.5 - 2 hours |
| Level | Intermediate |
| Prerequisites | GitHub account, basic Python knowledge |
What You'll Learn#
By the end of this workshop, you will be able to:
| Skill | Description |
|---|---|
| Instrument AI Applications | Add OpenLLMetry/Traceloop to Python AI apps |
| Visualize LLM Traces | See prompts, completions, and token usage |
| Analyze RAG Pipelines | Debug with distributed tracing |
| Use Dynatrace MCP | Query observability data from your IDE |
| Automate Workflows | Build AI cost alerts and daily summaries |
Workshop Agenda#
| Time | Lab | Description |
|---|---|---|
| 15 min | Lab 0: Environment Setup | Configure your GitHub Codespace |
| 15 min | Lab 1: AI Instrumentation | Add OpenLLMetry to the sample app |
| 30 min | Lab 2: Explore Traces | Analyze AI traces in Dynatrace |
| 30 min | Lab 3: Dynatrace MCP | Use MCP for agentic AI |
| 30 min | Lab 4: Workflow Automation | Automate AI cost monitoring |
What's Included#
- Pre-configured GitHub Codespace with all dependencies
- Sample RAG/LLM application (FastAPI + LangChain + ChromaDB + Azure OpenAI)
- Access to Dynatrace playground environment (instructor-provided)
- Step-by-step lab guides
Ready to Begin?#
Each attendee gets their own isolated Codespace. Your changes stay local to your Codespace and won't affect other attendees.
Need Help?#
- Raise your hand in the workshop
- Check the Resources page for documentation links
- Ask your workshop instructor