Introduction
Welcome to Aegis
Aegis is an evaluation platform for AI systems. The evaluation methodology walks a team from getting data in through choosing metrics, baselining, and gating on results. Use the API to run evaluations on your data, retrieve results and aggregates, export reports, and organize datasets.
Find more information about:
Evaluation Methodology
A practical guide for adopting Aegis: model your data, choose metrics and thresholds, baseline in development, automate evaluation, and gate on results.
API Documentation
Use the HTTP API at https://api.aegisevals.ai to run evaluations, manage dataset types, models, datasets, and collections, and export results—authentication, routes, and payloads are documented here.
Metrics Documentation
Metrics are organized by category—each overview lists the metrics available and how scores are computed.
About Aegis
Aegis provides evaluation infrastructure for AI systems—runs, metrics, and reporting so teams can measure quality, safety, and correctness at scale. Visit the main site for product and contact details.