DeepGraph

DeepGraph is a Next.js app that turns long-form research text into an interactive knowledge graph. You can upload a PDF or paste raw text, generate connected concepts with Gemini, and ask follow-up questions in an in-app chat.

Features

• Generate graph nodes and links from research text.
• Parse PDF files in the browser and extract text automatically.
• Explore relationships through an interactive D3 force-directed graph.
• Ask graph-grounded questions in the built-in chat panel.
• Click nodes for short AI-generated summaries.

Tech Stack

• Next.js 14 (App Router) + React 18 + TypeScript
• Tailwind CSS
• D3.js (graph rendering)
• PDF.js (PDF extraction)
• Gemini API (gemini-2.0-flash)

Requirements

• Node.js 18+
• npm 9+
• A Gemini API key

Quick Start

npm install
npm run dev

Open http://localhost:3000.

How To Use

1. Click Get Started.
2. Enter your Gemini API key.
3. Upload a PDF or paste report text.
4. Click Generate to build the graph.
5. Use the chat button to ask questions about the generated graph.
6. Click graph nodes to view short summaries.

Project Structure

• app/page.tsx: Main UI shell and app sections.
• public/app.js: Client-side logic for graph generation, D3 rendering, PDF parsing, and chat.
• app/globals.css: Global styles and graph/chat styling.
• vercel.json: Deployment headers and policy configuration.

Scripts

• npm run dev: Start development server.
• npm run build: Build for production.
• npm start: Run production server.
• npm run lint: Run lint checks.

Security Notes

• Your API key is entered in the browser UI and sent directly to the Gemini API from the client.
• Do not use high-privilege keys in shared/public environments.
• This project currently has no backend key proxy.

Deployment

Deploy on Vercel or any platform that supports Next.js.
If needed, adjust headers and policy settings in vercel.json.

License

This project is licensed under the MIT License. See LICENSE.