🚀 Next.js RAG Chatbot (Mistral AI)

A Retrieval-Augmented Generation (RAG) Chatbot built with Next.js, LangChain.js, Redis, and Mistral AI.
This chatbot enhances responses by retrieving relevant contextual data before generating answers.

Publication link : Publication

📂 Project Structure

📁 Chat_bot
│-- 📂 src
│   ├── 📂 app
│   │   ├── 📂 [...url]     # Dynamic route for chat sessions
│   │   ├── 📂 api          # API routes
│   │   ├── 📜 layout.tsx
│   │   ├── 📜 page.tsx
│   ├── 📂 components       # Reusable React components
│   ├── 📂 lib              # Utility functions (RAG, Redis)
│   ├── 📂 styles           # Styling (CSS, Tailwind, etc.)
│-- .gitignore
│-- package.json
│-- next.config.js
│-- README.md

⚙️ Installation & Setup

1️⃣ Clone the repository

git clone https://github.com/kunal534/Chat_bot.git 
cd YOUR_REPO

2️⃣ Install dependencies

npm install

3️⃣ Set up environment variables

Create a .env.local file and add:

MISTRAL_API_KEY=your_mistral_api_key
REDIS_URL=your_redis_url
NEXT_PUBLIC_API_BASE_URL=http://localhost:3000/api

4️⃣ Run the development server

npm run dev

Your chatbot should now be running at http://localhost:3000 🎉

---

🛠️ Tech Stack

• Frontend: Next.js, React
• Backend: Next.js API routes
• AI Model: Mistral AI for response generation
• Database: Redis for caching and session management
• Retrieval Layer: LangChain.js (RAG pipeline)

---

🔧 Features

✅ Supports RAG-based responses
✅ Stores chat history in Redis
✅ Handles multi-session conversations
✅ Supports dynamic URL-based contexts
✅ Easy to deploy on Vercel or Railway

---

🚀 Deployment

Deploy to Vercel

vercel

Future Scope

Document Content Restriction Handling:

Digital Rights Management (DRM)
Content Security Policy (CSP)
Other access-control restrictions
In such cases, the RAG pipeline cannot generate embeddings from the content, which limits contextual accuracy.

Potential Enhancements:

Integrating OCR-based content extraction for protected documents (where legally permissible)
Expanding parsers to handle non-standard file formats with embedded restrictions
Support for proxy retrieval through secure, authenticated sessions