Skip to content

Harshitazz/Vectra-mind

Repository files navigation

AI-Powered Multilingual Document Query System

Live Demo

Overview

This system is designed to provide accurate answers to queries from uploaded documents or URLs. It leverages LangChain, FAISS, Hugging Face Sentence Transformers, and Large Language Models (LLMs) to extract and store information as vector embeddings, enabling efficient semantic search with multilingual support.

Features

  • Document-based Question Answering
  • Multilingual Semantic Search
  • Supports PDF Uploads and URL Extraction
  • Efficient Vector Search with FAISS
  • Optimized LLM Calls for Reduced Latency
  • Secure API with Rate Limiting
  • FastAPI Backend Deployed on AWS

Tech Stack

  • Backend: FastAPI
  • Vector Database: FAISS
  • Machine Learning: Hugging Face Sentence Transformers, LangChain, LLMs
  • Deployment: AWS

Getting Started

Enjoy using the AI-Powered Multilingual Document Query System! 🚀

About

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages