BetterRAG: Powerful RAG evaluation toolkit for LLMs. Measure, analyze, and optimize how your AI processes text chunks with precision metrics. Perfect for RAG systems, document processing, and embedding quality assessment.

An LLM-RAG powered application where you can talk with your own pdf.

AI-powered semantic matching system that leverages local embedding technology (via @xenova/transformers) and Pinecone's vector database to perform intelligent matching based on skills, experience, and job requirements.

cython based chunking package

A RAG-powered knowledge assistant that answers style guide questions for translators. Built with Weaviate, Amazon Bedrock (Claude 3 Sonnet), and Streamlit.

A deep dive into text chunking for Retrieval-Augmented Generation systems