ContextQMD
Back to Nexa Sdk

RAG Pipeline Demo

cookbook/PC/RAG-LLM/README.md

0.2.73917 B
Original Source

RAG Pipeline Demo

About

This project is a demo for building a Retrieval-Augmented Generation (RAG) pipeline using NexaSDK and NexaAI Python binding. It showcases how to combine state-of-the-art embeddings, reranking, and generation models to answer questions over your own documents.

Key features:

  • ๐ŸŒ Multi-platform support โ€” Works on Snapdragon NPU (Windows ARM), macOS, and Windows x64
  • ๐Ÿ”„ End-to-end RAG demo โ€” From document ingestion to retrieval and answer generation
  • ๐Ÿ’ป Local execution โ€” All processing happens on your device; no data leaves your machine
  • โšก Easy to run โ€” Minimal setup to explore NexaSDK / NexaAI capabilities

Bring your own files (PDFs, Word docs, text) and ask questionsโ€”the system retrieves relevant context and generates answers entirely on your device.

Examples