KnowGenie

A retrieval-augmented generation (RAG) application that enables natural language querying over uploaded documents. Built with Gemini 2.5 Flash, FAISS vector storage, FastAPI, and Next.js.

---

Architecture

knowgenie/
├── backend/
│   ├── server.py             # FastAPI REST API
│   ├── vectors.py            # Document loading, chunking, FAISS indexing
│   ├── chatbot.py            # Gemini 2.5 RAG retrieval chain
│   ├── app.py                # Streamlit interface (standalone alternative)
│   └── requirements.txt
│
└── frontend/
    ├── app/                  # Next.js app router
    ├── components/
    │   ├── Landing.tsx       # API key entry screen
    │   ├── AppShell.tsx      # Layout
    │   ├── Sidebar.tsx       # File upload and document list
    │   └── ChatPanel.tsx     # Chat interface
    └── lib/api.ts            # Backend API calls

Stack: Python 3.10+, FastAPI, LangChain, Google Generative AI, FAISS, Next.js 14, Tailwind CSS, Framer Motion

---

Prerequisites

• Python 3.10 or higher
• Node.js 18 or higher
• A Google AI Studio API key — get one free

---

Installation

Backend

cd backend
python -m venv venv

# Windows
venv\Scripts\activate

# macOS / Linux
source venv/bin/activate

pip install -r requirements.txt

Frontend

cd frontend
npm install

---

Running the Application

Two terminal processes are required.

Terminal 1 — API server

cd backend
uvicorn server:app --reload --port 8000

Terminal 2 — Frontend

cd frontend
npm run dev

Open http://localhost:3000 in your browser.

---

Usage

1. Enter your Google API key on the landing screen.
2. Upload a document (PDF, TXT, or CSV) using the sidebar.
3. Wait for the embedding process to complete. Progress is displayed in real time.
4. Ask questions about the document in the chat interface.

---

API Reference

Method	Endpoint	Description
GET	/health	Health check
POST	/upload	Upload a document
POST	/embed	Start background embedding
GET	/embed/status/{filename}	Poll embedding status
POST	/chat	Submit a query
GET	/documents	List uploaded documents
DELETE	/documents/{id}	Remove a document

Interactive API documentation is available at http://localhost:8000/docs.

---

Supported File Types

Format	Loader	Chunking Strategy
PDF	PyPDFLoader	RecursiveCharacterTextSplitter (800 tokens, 100 overlap)
TXT	TextLoader	RecursiveCharacterTextSplitter (800 tokens, 100 overlap)
CSV	CSVLoader	One document per row

---

Rate Limits

The free tier of the Gemini Embedding API allows 100 requests per minute. The pipeline batches chunks in groups of 8 with a 0.7-second delay between batches, keeping throughput at approximately 68 RPM. For larger files or higher throughput, enabling billing on Google AI Studio removes this constraint.

---

Troubleshooting

Embedding fails with a 429 error The free tier rate limit has been reached. The system will automatically retry after a 20-second cooldown. If failures persist, reduce BATCH_SIZE in vectors.py or enable billing.

"Failed to fetch" on upload Confirm the backend is running on port 8000 and that npm run dev was restarted after any changes to next.config.js.

Chatbot returns no results The FAISS index must exist before the chatbot loads. Ensure embedding completed successfully before submitting a query.

Port conflict Change the port in the uvicorn command and update the proxy destination in frontend/next.config.js accordingly.

---

Alternative Interface

The original Streamlit interface remains functional as a standalone option:

cd backend
streamlit run app.py

Available at http://localhost:8501.

---

License

MIT