Ever wondered how to keep an eye on your super smart AI applications once they're out in the wild? Carly Richmond's talk from NDC Manchester spills the beans on observing production LLM and GenAI apps. Forget just logs; she stresses the need for tracing and metrics to truly understand how these systems are behaving and where things might be going wrong.
She even shows you the ropes with a hands-on example, instrumenting a TypeScript AI agent using OpenLit and OpenTelemetry. Get ready to grab those crucial signals to detect and fix common issues in your GenAI setups!
Watch on YouTube
Top comments (0)