ModelReceipt

ModelReceipt is a web app (with optional desktop agent) that turns messy AI development artifacts into audit-ready “evidence packs” for AI governance and safety reviews. It continuously collects model cards, dataset lineage, eval results, prompt templates, red-team findings, incident logs, and approval trails from your existing tools (GitHub, Jira, MLflow, cloud logs). Then it maps them to common frameworks (EU AI Act, NIST AI RMF, ISO/IEC 23894) and generates a structured, timestamped compliance dossier you can hand to legal, procurement, or an external auditor. This is not a magic ‘certify my model’ button: it’s a ruthless system of record that highlights missing evidence, inconsistent claims, and untested risk areas. Teams get a dashboard showing what’s compliant, what’s unverifiable, and what will fail an audit—before regulators or customers do.