Skip to content

Srijan3412/DocuMind-AI

Repository files navigation

DocIntel AI — Raw Papers to Structured Signals

DocIntel AI (formerly DocuMind AI) is an industrial-grade document intelligence platform and visual dashboard. It allows you to process complex documents (like invoices, resumes, and legal papers) using a scalable, async AI pipeline.

This repository contains the client-side Single Page Application (SPA) built using React, Vite, Tailwind CSS v4, and TypeScript.

Features

  • Drag-and-Drop Dropzone: Sleek interface supporting PDF, TXT, and Markdown files up to 20MB.
  • Async Processing Pipeline: Real-time progress visualization showing layout recognition, Gemini AI analysis, and structured JSON generation stages.
  • Interactive JSON Viewer: Displays live-parsed structured signals extracted from documents.
  • Visual Intelligence Console: Deep analytics panel showing document summaries, key metadata (document type, confidence score, language), extracted entities (with confidence levels), and full details.
  • Local History Tracking: Document records and analysis history are saved client-side using localStorage.
  • Client-Side PDF Parsing: Robust offline text extraction using pdfjs-dist worker threads.

Architecture

graph TD
  UI[React + Vite SPA] -->|Upload File| API[Express API Gateway]
  API -->|Push Job| PubSub[Google Cloud Pub/Sub]
  PubSub -->|Pull Job| Worker[Cloud Run Worker]
  Worker -->|Parse PDF & Call Gemini| Gemini[Gemini 2.5 Flash]
  Worker -->|Store Result| DB[(Firestore Database)]
  UI -->|Poll Status| API
  API -->|Fetch Status| DB
Loading

Tech Stack

  • Core: React 19, TypeScript 5, Vite 7
  • Styling: Tailwind CSS v4 (using the @tailwindcss/vite compiler plugin)
  • Routing: React Router DOM (Vite SPA mode)
  • Query Management: TanStack React Query v5
  • UI Components: Radix UI primitives & Lucide React icons
  • Notifications: Sonner

Getting Started

Prerequisites

  • Node.js (v18 or higher recommended)
  • npm or bun

Installation

  1. Install dependencies:

    npm install
  2. Start the local development server:

    npm run dev

    The app will run locally at http://localhost:8080/.

Production Build

To build the static application assets:

npm run build

This outputs minified static assets in the /dist directory, ready to be hosted on static hosting services (like Vercel, Netlify, or GitHub Pages).


Deployment on Vercel

The application is fully configured as a Single Page Application for Vercel.

  • Routes are rewritten client-side via vercel.json to ensure routing handles index.html fallback correctly without 404s.
  • Environment variables should be configured directly in your Vercel Project Settings if your backend requires them.

About

AI-powered document intelligence platform for structured extraction, summarization, and workflow automation using Gemini, Cloud Run, and Pub/Sub.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages