Core concepts overview

This section breaks down how LevelApp works behind the scenes.

You’ll learn about the core concepts that power the system:

1. Simulators – how conversations are triggered and responses collected
2. Evaluators – how replies are scored against expectations
3. Scoring – how similarity is measured and results are aggregated
4. Batch Tests – how evaluations are structured and submitted

Each piece works together to help you test AI models reliably and consistently.

Tip: If you’re just exploring, start with Simulators to see how evaluations begin.