Dingo

Dingo is a comprehensive data quality inspection toolkit designed for AI and data teams who need to ensure high-quality datasets and reliable AI models. It combines rule-based checks, LLM-based evaluation, and Agent-as-a-judge modes, and supports both quick sample validation and full-scale batch evaluation. Dingo offers robust, scalable inspection (including text and image data), hallucination detection for RAG pipelines, visual health check reports, and integrates via CLI, SDK, Gradio, and as an MCP Server into broader workflows.