Welcome to my full stack chatbot application powered by the Gemini API! This project allows users to:
✅ Upload and chat with PDF documents
✅ View and store chat history
✅ Authenticate securely via Supabase
✅ Interact through a polished React + Next.js UI
✅ Experience a persistent session with real-time messaging
- 🔐 User Authentication using Supabase (Signup, Login, Logout)
- 📄 PDF Upload with automatic text extraction using
pdf-parse - 🤖 Chat with AI via Gemini API integration
- 💬 Chat History stored and retrieved per user
- 📁 Persistent Sessions across multiple logins
- 🧾 Database-backed with PostgreSQL for chats/messages
- 🧠 Intelligent Context Handling for more relevant responses
- 🎯 Fully responsive and styled with TailwindCSS
- 🚀 Deployed via Vercel
| Frontend | Backend/API | Auth & DB |
|---|---|---|
| Next.js 14 | Next.js App Router | Supabase (PostgreSQL) |
| React | REST API Routes | Supabase Auth (JWT) |
| TailwindCSS | Gemini API | Prisma-like Supabase queries |
- Authentication: Supabase manages user sessions via JWT tokens.
- PDF Upload: Files are parsed using
pdf-parseto extract text on the server. - Gemini AI: User queries + extracted content are sent to Gemini for response generation.
- Chat History: Chats are saved in a Supabase database under each user.
- Retrieval: Users can open old conversations via a chat list UI.
📁 app/ ├── api/ → Custom API routes for Gemini & PDF ├── auth/ → Login & signup pages ├── chat/ → Main protected chat page ├── components/ → UI components (ChatUI, ChatList, Header, etc.) ├── lib/ → Business logic (Gemini, DB queries) ├── utils/ → Supabase clients 📁 public/ → Static assets (SVGs) 📁 docs/ → Screenshots, schema, sample responses
See full structure here
File: docs/sample_chat.txt
User: hello im texting and making documentation with u say hi
AI: Hi! Happy to help with your texting and documentation. Let me know how I can assist you!
User: this is for my intern u know ?
AI: Okay, I understand this is for your intern. How can I help them (or you) with their tasks or documentation?
User: and i have made u in a way u can remember chats and load them and have history too. with auth isnt it cool?
AI: That's really cool! You've created a system with chat memory, history loading, and authentication. Awesome work!
🛠️ Setup Instructions
Clone the repo
git clone https://github.com/your-username/gemini-chatbot.git
cd gemini-chatbot
Install dependencies
npm install
Configure Supabase & API Keys
Add a .env.local file:
NEXT_PUBLIC_SUPABASE_URL=your-supabase-url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your-anon-key
GEMINI_API_KEY=your-gemini-api-key
Run locally
npm run dev
🗃️ Database Schema
See docs/database_schema.txt for table structures:
users
chats
messages
📤 Deployment
✅ Hosted on Vercel
(Just connect your GitHub repo and set environment variables!)
🧾 Deliverables
✅ Source Code (frontend, backend, and DB scripts)
✅ Sample Responses in .txt under /docs
✅ Deployment Link to live app
✅ Well-documented code with comments and clean structure
🙌 Credits
Supabase for auth and DB
Gemini API for chatbot
pdf-parse for PDF extraction
Made with ❤️ for Internship Assignment
---
Let me know if you want this exported as a `.md` or `.pdf` file or edited further for submission!