📄 PDF RAG System

Upload PDFs and query them using an open-source LLM (Mistral via Ollama).

---

🧰 Requirements

Install:

• Node.js (>= 18)
• Python (>= 3.9)
• pip

---

⚙️ 1. Install & Run Ollama (LLM)

Install

Mac

brew install ollama

Linux

curl -fsSL https://ollama.com/install.sh | sh

Windows

• Download from: https://ollama.com/download
• Install the .exe
• It will run as a background service

---

Start Ollama

Mac/Linux

ollama serve

Windows

• Ollama usually starts automatically after install
• If not, open Command Prompt / PowerShell:

ollama serve

---

Download model

ollama pull mistral

(Optional test)

ollama run mistral

---

🖥️ 2. Backend Setup (FastAPI)

cd backend
python -m venv .venv

Activate virtual environment

Mac/Linux

source .venv/bin/activate

Windows (CMD)

.venv\Scripts\activate

Windows (PowerShell)

.venv\Scripts\Activate.ps1

---

Install dependencies

pip install fastapi uvicorn chromadb pymupdf sentence-transformers requests

---

Run backend

uvicorn main:app --reload

Runs at:

http://localhost:8000

---

💻 3. Frontend Setup (React)

cd frontend
npm install
npm run dev

Runs at:

http://localhost:5173

---

▶️ Run Everything (order matters)

1. Start Ollama

ollama serve

2. Start backend

cd backend
uvicorn main:app --reload

3. Start frontend

cd frontend
npm run dev

---

🧪 Usage

1. Open frontend in browser

2. Upload a PDF

3. Ask questions like:

   • "What is the total amount?"
   • "Who is billed?"

---

⚠️ Common Issues

LLM not responding

ollama serve

First run is slow

• Mistral downloads ~4–5GB (one-time)

PowerShell script blocked (Windows) Run:

Set-ExecutionPolicy RemoteSigned -Scope CurrentUser

Empty answers

• PDF may not contain extractable text (scanned PDFs need OCR)

---

📌 Notes

• Uses ChromaDB (local vector DB)
• Uses sentence-transformers for embeddings
• If answer not found → returns:

Not found in document

---

✅ Done

If all 3 services are running (Ollama → Backend → Frontend), the system should work end-to-end.