/Ask-It

From Static to Dynamic: Transforming PDFs into Interactive Conversations with LLMs

Primary LanguageJavaScript

Ask-It

Talk with your PDFs using Retrieval Augmented Generation (RAG) and LLMs

Overview

This repository contains a Web application built using Django Rest framework and React. The application allows users to upload PDF(s), ask questions related to the PDF(s) and get accurate RAG answers using LLMs.

Features

  • User Authentication
  • Create multiple conversations
  • Ask questions based on the uploaded PDF and get accurate results
  • User friendly interface

Tech Stack

FrontEnd

  • React (UI)
  • Tailwind CSS
  • Parcel (Bundler)
  • Nginx (Server)

BackEnd

  • Django Restframework (APIs)
  • JSON Web Token (JWT) Authenticatoin
  • PostgreSQL with PG Vector (Vector DB)
  • Sentence transformers (Embeddings)
  • Google FLAN-T5 (LLM)
  • Hugging Face Hub
  • Langchain
  • Gunicorn (Server)
  • Nginx (Reverse Proxy)

Screenshots

LOGIN

Screenshot from 2023-10-16 19-44-33

SIGN UP

Screenshot from 2023-10-16 19-44-30

Demo

Ask.It.demo.webm

Know more about how it works Blog