# Who This Is For

TrustLens is designed for machine learning practitioners who need to move beyond simple performance metrics to build truly reliable systems.

## ML Engineers
Building mission-critical production systems where a high-confidence mistake has real-world consequences. Use TrustLens in your CI/CD pipeline to catch regressions in calibration or bias.

## Data Scientists
When you need to justify model decisions to stakeholders or regulators. TrustLens provides the visual and quantitative evidence needed to explain model behavior.

## Researchers
Benchmarking the reliability of new architectures. Go beyond the leaderboard accuracy and compare how different models represent classes or handle difficult edge cases.

## AI Governance & Compliance Teams
Focused on safety, fairness, and regulatory compliance (like the EU AI Act). Subgroup performance and bias reports provide the transparency required for auditing.

---

## Prerequisites
TrustLens is designed to be **zero-friction**. If you have a model with a `.predict()` and `.predict_proba()` method (like `scikit-learn`, `XGBoost`, `LightGBM`), you're ready to go.