Built the first version of AI Agent Hallucination Tester!
The platform automatically tests AI-generated invoice data against ground-truth records, detects hallucinations, calculates risk scores, and routes high-risk cases for human review.
Features completed: • AI-powered invoice validation using Google Gemini • Hallucination detection engine • Risk scoring system (0–100) • Interactive analytics dashboard • Human review workflow • Detailed reporting and insights
Currently working on deeper UiPath integration and expanding support for additional business document workflows.
Log in or sign up for Devpost to join the conversation.