ModelReceipt

ModelReceipt is a web app that generates “audit-ready” evidence packs for AI features without forcing teams into heavyweight governance programs. It connects to your model endpoints, data sources, and CI/CD to automatically capture what regulators and enterprise buyers actually ask for: model versioning, training data lineage summaries, evaluation results, known limitations, human oversight steps, and incident logs. It then produces a standardized, timestamped report (plus a shareable link) you can attach to procurement questionnaires, SOC2-style reviews, or internal risk committees. This is not a magic fairness detector; it’s a practical compliance workflow that reduces scramble and spreadsheet chaos. The app includes lightweight checklists aligned to common frameworks (EU AI Act-style risk tiers, NIST AI RMF), and it flags missing artifacts rather than pretending to judge ethics perfectly.