Regrada

Regrada is a CI gate for LLM behavior that catches behavioral regressions in AI applications before they reach production. It records live LLM traffic via an HTTP proxy, converts traces into version-controlled test cases, and enforces policies in CI with zero code changes. It targets developers, IT, and executives in B2B companies, from startups to enterprises. It is positioned as a CI/CD infrastructure for AI systems, differentiating through its proxy-based traffic capture and policy enforcement in CI.

Key features

• Zero-code traffic capture via HTTP proxy
• Supported providers: OpenAI, Anthropic, Azure OpenAI, AWS Bedrock
• Intelligent policy engine with assertions and variance thresholds
• Automatic PII and secrets redaction
• Native GitHub Actions integration with PR comments
• Web dashboard for trace visualization and comparison