Monitor End-to-End Agentic AI and LLMs with Dynatrace AI Observability

Explore Dynatrace's AI Observability App in this 11-minute video tutorial that demonstrates comprehensive monitoring and analysis capabilities for Generative AI, Agentic AI, and LLM services. Learn to instrument applications using OpenLLMetry, set up monitoring through the configuration wizard, and leverage the service health view to track agents, models, and key performance metrics. Discover how to create intelligent alerts, visualize AI service interactions through the explorer view, and trace complex agentic workflows while examining detailed prompt information. Master exception analysis techniques, utilize pre-built dashboards for popular providers like OpenAI and Amazon Bedrock, and implement prompt audit trails for model comparison. Gain insights into cost analytics and optimization features that help manage expenses across multiple AI providers, ensuring your AI-powered applications maintain reliability, efficiency, and transparency throughout their operational lifecycle.

Syllabus

– Introduction: What is AI Observability?
– Overview of the App and Key Capabilities
– Getting Data In: OpenLLMetry and Setup Wizard
– Service Health View: Agents, Models and Metrics
– Creating Alerts
– Explorer View: Visualizing AI Service Interactions
– Tracing Agentic Workflows and Prompt Details
– Analyzing Exceptions
– OpenAI, Amazon Bedrock, and other ready-made dashboards
– Prompt Audit Trail and Model Comparison
– Cost Analytics and Optimization Features
– Wrap-Up: Why AI Observability Matters