Playbook Overview

Welcome to Confident AI's evaluation playbook! This playbook covers our most commonly asked questions from companies working with us to build high-quality AI apps through evals and observability. It also contains a collection of eval advice that we've (Kritin and Jeffrey) accumulated over the past 3 years of building Confident AI.

This playbook is a good starting point for those with questions about best practices and workflows in general. If you're already working with the Confident AI team, feel free to reach out to your dedicated support channel.

This playbook also assumes that you are using, or are about to use, Confident AI. If you are using another LLM evaluation and observability tool, you're still free to read this. Simply mentally swap out Confident AI with whatever platform you're using, and if they don't support the features and workflows we talk about here, that's a great sign that you should consider switching 😊