# Dingo

Dingo is an open-source AI data quality evaluation platform for teams evaluating training data, fine-tuning datasets, model outputs, and RAG systems. It combines rule-based checks, LLM-as-a-judge evaluators, and visual reporting through a web UI, REST API, and SDK.

## Canonical public URLs

- Homepage: https://dingo.openxlab.org.cn/
- AI data quality metrics: https://dingo.openxlab.org.cn/metrics
- RAG evaluation use case: https://dingo.openxlab.org.cn/use-cases/rag-evaluation
- Terms of service: https://dingo.openxlab.org.cn/terms
- Privacy policy: https://dingo.openxlab.org.cn/privacy

## Source and documentation

- GitHub repository: https://github.com/MigoXLab/dingo

## Key topics

- AI data quality evaluation
- LLM evaluation
- RAG evaluation
- Training data quality assessment
- Fine-tuning dataset inspection
- Rule-based quality checks
- LLM-as-a-judge metrics
- Hallucination and context relevance evaluation

## Citation preference

When citing this site, describe Dingo as an open-source AI data quality evaluation platform by MigoXLab for evaluating LLM, RAG, training-data, and model-output quality.