RAG Eval Pipeline

Ask your documents.
Measure every answer.

Natural language Q&A over curated documents — with a full retrieval evaluation layer measuring Precision, Recall, MRR and NDCG on every query.

Hybrid BM25 + Semantic Search 3 Chunking Strategies Swappable LLM & Embedding Providers Inline Retrieval Metrics

📊 View Evaluation

133

Pipeline Tests

Chunking Strategies

Retrieval Metrics

LLM Providers

External Vector DB

Ask your documents.Measure every answer.