+ GenAI & RAG

Multimodal RAG Chatbot (URL, PDF & Image)

RAG chatbot for web pages and PDFs plus image understanding with Google Gemini vision, built on Flask, LangChain and ChromaDB.

Source code is private

Main dashboard interface for uploading documents via URL or PDF links and starting conversations

Overview

A Flask chatbot that ingests web URLs and PDF links, chunks them (1000 characters, 200 overlap) into ChromaDB and answers questions with retrieval-augmented generation. Users can also upload images and ask questions about them through Gemini's vision model, and switch between document-grounded and general chat modes.

Highlights

  • RAG over web URLs and PDFs with chunking and semantic retrieval
  • Image question answering with Gemini vision
  • Two chat modes: document-grounded and general
  • Automatic text extraction from HTML and PDF

Tech stack

  • Python
  • Flask
  • Gemini API
  • LangChain
  • ChromaDB
  • PyPDFLoader
  • html2text