🤖 RAG Chatbot — Acme Tech Solutions

A production-ready Retrieval-Augmented Generation (RAG) Chatbot built with:

• 🐍 Backend — FastAPI + LangChain + Pinecone (vector search) + SQLite (chat history) + OpenAI
• ⚛️ Frontend — React 19 + Vite + Tailwind CSS

---

📁 Project Structure

RAG_Chatbot/
├── rag_chatbot/          # Python FastAPI backend
│   ├── routes/           # API route handlers (chat, history, session, ingest)
│   ├── middleware/       # Auth middleware
│   ├── docs/             # Source documents for RAG ingestion
│   ├── main.py           # FastAPI app entry point
│   ├── config.py         # Configuration (reads .env)
│   ├── vectorstore.py    # Pinecone vector store logic + local BM25 fallback
│   ├── embeddings.py     # OpenAI embedding setup
│   ├── memory.py         # SQLite chat session management
│   ├── ingest_docs.py    # One-shot script to load docs into Pinecone
│   ├── requirements.txt  # Python dependencies
│   └── .env.example      # Backend env template
│
├── rag_frontend/         # React + Vite frontend
│   ├── src/
│   │   ├── components/   # ChatWindow, Sidebar, MessageBubble
│   │   ├── App.jsx       # Root component
│   │   ├── api.js        # Axios API calls to backend
│   │   └── index.css     # Global styles
│   ├── package.json
│   └── .env.example      # Frontend env template
│
├── .gitignore
└── README.md

---

✅ Prerequisites

Make sure you have the following installed:

Tool	Version	Download
Python	≥ 3.10	python.org
Node.js	≥ 18.x	nodejs.org
Git	any	git-scm.com

You will also need accounts / API keys for:

• Pinecone — pinecone.io (free tier works)
• OpenAI — platform.openai.com (needs an API Key with credits)

---

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/Priyap1038/ChatBot.git
cd ChatBot

---

2. Backend Setup (rag_chatbot/)

a) Create a virtual environment

cd rag_chatbot

# Windows
python -m venv venv
venv\Scripts\activate

# macOS / Linux
python3 -m venv venv
source venv/bin/activate

b) Install dependencies

pip install -r requirements.txt

c) Configure environment variables

# Copy the example file
copy .env.example .env        # Windows
cp .env.example .env          # macOS / Linux

Now open .env and fill in your values:

OPENAI_API_KEY=sk-...                   # Your OpenAI API key
PINECONE_API_KEY=pcsk_...              # Your Pinecone API key
PINECONE_INDEX_NAME=priya-rag-index    # Your Pinecone index name
CORS_ORIGINS=*                         # Use * for local dev
RATE_LIMIT=30/minute
LOG_LEVEL=INFO

d) Create Pinecone Index

⚠️ CRITICAL: Create your index in Pinecone with Dimensions = 1536 and Metric = cosine. This matches the default OpenAI text-embedding-3-small model output.

e) Ingest documents into Pinecone

This step uploads your documents in docs/ into the Pinecone vector store and builds the local BM25 search corpus (bm25_corpus.json). Run this once before starting the server (or whenever you physically add new files to docs/). The local search state survives restarts!

python ingest_docs.py

f) Start the backend server

uvicorn main:app --reload --port 8000

The API will be live at: http://localhost:8000

• Swagger UI: http://localhost:8000/docs
• Health check: http://localhost:8000/api/health

---

3. Frontend Setup (rag_frontend/)

Open a new terminal and run:

a) Install Node dependencies

cd rag_frontend
npm install

b) Configure environment variables

# Windows
copy .env.example .env

# macOS / Linux
cp .env.example .env

For local development, the default .env values work out of the box (Vite proxies /api/* → localhost:8000):

VITE_API_URL=       # Leave empty for local dev
VITE_API_KEY=       # Leave empty unless backend API_KEY is set

c) Start the frontend dev server

npm run dev

The app will be live at: http://localhost:5173

---

🖥️ Running Both Together (Quick Start)

Open two terminals side-by-side:

Terminal 1 — Backend	Terminal 2 — Frontend
cd rag_chatbot	cd rag_frontend
venv\Scripts\activate	npm install
uvicorn main:app --reload	npm run dev

Then open http://localhost:5173 in your browser. 🎉

---

📄 Adding Your Own Documents

1. Place your .md, .txt, or .pdf files inside rag_chatbot/docs/
2. Re-run the ingestion script:

   cd rag_chatbot
   python ingest_docs.py

3. Restart the backend server.

---

🔌 API Endpoints

Method	Endpoint	Description
GET	/api/health	Health check
GET	/api/sessions	List all chat sessions
POST	/api/chat	Send a message
GET	/api/history/{session_id}	Get chat history for a session
POST	/api/session	Create a new session
POST	/api/ingest	Ingest a document via API

Full interactive docs: http://localhost:8000/docs

---

🏗️ Production Build (Frontend)

cd rag_frontend
npm run build

Output will be in rag_frontend/dist/. Serve it with any static host (Vercel, Netlify, etc.).

For the backend, update .env:

CORS_ORIGINS=https://yourfrontend.com
API_KEY=your-strong-secret-key

---

🛠️ Troubleshooting

Problem	Fix
ModuleNotFoundError	Make sure your venv is activated and pip install -r requirements.txt was run
Pinecone Dimension mismatch error	Delete the index and recreate it with dimension = 1536
OpenAI 401 / Quota Exceeded	Check your OpenAI billing page to ensure your key has credits
Empty responses / no context	Run python ingest_docs.py to populate Pinecone and the local database
CORS errors in browser	Ensure CORS_ORIGINS=* is set in backend .env during development
Frontend can't reach backend	Make sure backend is running on port 8000 and frontend on 5173

---

📜 License

This project is for educational / internal use. Feel free to fork and adapt!