Observing AI Applications with OpenLit and OpenTelemetry

Learn to implement comprehensive observability for production AI applications using OpenLit and OpenTelemetry in this conference talk from NDC Manchester. Explore the critical need for monitoring LLMs and GenAI applications beyond traditional logging, focusing on tracing and metrics that provide insights into usage patterns and error handling within AI-powered systems. Discover best practices for observing production applications that leverage large language models, with practical demonstrations of instrumenting an AI agent application built in TypeScript. Examine the specific signals and data points essential for identifying and resolving common issues that arise in production GenAI environments, gaining hands-on knowledge of tools and techniques that enable effective monitoring and troubleshooting of AI-driven applications at scale.