VeritasRAG
What it is
VeritasRAG is a multi-agent Retrieval-Augmented Generation system built with LangChain and LangGraph that uses a stateful reflection loop to verify answers. It expands queries, retrieves documents from ChromaDB, reranks them with a cross-encoder, and iteratively refines the generated response until a Fact-Checker Agent confirms it is supported by the retrieved evidence.
The problem it solves
Traditional RAG pipelines often produce hallucinated answers because they perform a single retrieval-generation pass without validating the output against the source context.
What you could use it for
- Reducing hallucinations in technical documentation Q&A
- Building self-correcting RAG pipelines for local LLMs
- Interactive document querying with confidence scoring via Gradio
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.