Glossary

Online Evaluation

An evaluation performed on live or production-like interactions to continuously assess an AI system's behavior, quality, safety, or performance.

Explore more about Evaluation Methods

Related terms

Model Monitoring

The practice of continuously tracking an AI model's performance, behavior, usage, and operational health in production.

Continuous Evaluation

An evaluation approach that continuously measures AI system performance throughout development and production to detect regressions and ensure quality.

Signal, not noise.

Focused newsletter for builders and knowledge workers tracking how AI is changing real work. We surface what matters in practice, not every headline. Curated for practitioners, not spectators.

Subscribe to our newsletter