Build Better AI Agents Workshop: LLM Evaluation & Observability with Opik
Learn the evaluation and observability techniques that separate experimental projects from production-ready systems directly from the Opik team.
About This Workshop
This session is part of Comet's global New Year Resolution AI Hackathon (but you can join even if you're not competing—and it's not too late to sign up!).
Abby Morgan will walk you through Opik, the open-source platform for LLM evaluation, observability, and optimization.
What You'll Learn:
Log your first trace and understand what's happening inside your LLM
Create datasets for systematic testing
Build custom evaluation metrics that matter for your use case
Evaluate programmatically and through Opik's online interface
Use the prompt playground to iterate faster
Leverage the agent optimizer for better performance
Unlock Opik Assist for AI-powered insights
Why This Matters
Whether you're building for the hackathon (where evaluation/monitoring is a judging criterion) or just want to ship more reliable AI systems, you'll leave with practical skills to measure and improve your agent's behavior. No more crossing your fingers and hoping your LLM does the right thing.
About the Hackathon
Comet is sponsoring a global hackathon challenging developers to build AI agents that help people stick to their New Year's resolutions. Join solo or with a team, learn from experts, and compete for prizes. [Link to register]
Speaker: Abby Morgan, Developer Advocate at Comet