COGNITO

Private. Local. Fast.

A fully local AI assistant that runs on your machine. No cloud, no data collection, no subscriptions. Your conversations stay yours.

---

✨ Features

🔒 100% Private

• All inference runs locally on your hardware
• No data ever leaves your machine
• No API keys or cloud accounts required

🚀 Fast & Lightweight

• Powered by llama.cpp for efficient inference
• Supports GGUF quantized models
• Works on Apple Silicon, NVIDIA GPUs, or CPU

🔍 Intelligent Search

• Built-in web search (model decides when to search)
• Deep search mode for comprehensive research
• Automatic context injection

📄 Document RAG

• Upload PDFs and text files
• Chat with your documents

🎨 Modern Interface

• Clean, dark-themed UI
• Chat history with session management
• System prompt customization

⬇️ In-App Model Downloads

• Browse models directly from Hugging Face
• One-click download of any GGUF model
• Automatic model management
• No external tools needed

---

🛠️ Installation

Prerequisites

1. Node.js (v18+)

• Check if installed:

  node --version

• Install:
  • Download Node.js (LTS recommended)
  • Or use a version manager like nvm (Mac/Linux) or nvm-windows.

2. Python (v3.10+)

• Check if installed:

  python --version  # or python3 --version

• Install:
  • Download Python
  • Windows Users: Make sure to check "Add Python to PATH" during installation.

3. GPU Acceleration (Optional but Recommended)

• macOS: Apple Silicon (M1/M2/M3) works out of the box.
• Windows/Linux: NVIDIA GPU with CUDA drivers installed.

Quick Start

1. Clone the repository

   git clone https://github.com/ArjunDeshwal/cognitoai.git
   cd cognitoai

2. Set up Python backend

   macOS/Linux:

   python3 -m venv venv
   source venv/bin/activate
   pip install -r backend/requirements.txt

   Windows:

   python -m venv venv
   venv\Scripts\activate
   pip install -r backend/requirements.txt

3. Install frontend dependencies

   cd app
   npm install

4. Run the app

   npm run electron:dev

5. Download a model

   • Click "⬇️ Models" in the header
   • Browse and download a GGUF model
   • Select it to start chatting!

---

📁 Project Structure

cognito/
├── app/                    # Electron + React frontend
│   ├── src/
│   │   ├── App.tsx         # Main React component
│   │   ├── App.css         # Styles
│   │   ├── services/api.ts # API client
│   │   └── components/     # UI components
│   ├── electron/
│   │   ├── main.cjs        # Electron main process
│   │   └── preload.cjs     # Preload script
│   └── package.json
├── backend/
│   ├── server.py           # FastAPI server
│   ├── document_rag.py     # Document processing
│   └── requirements.txt
├── tools/
│   └── info/web_search.py  # Web search tool
└── README.md

---

🤝 Contributing

Open source on GitHub ·

Development Setup

1. Fork the repository
2. Create a feature branch

   git checkout -b feature/amazing-feature

3. Make your changes
4. Test thoroughly
5. Commit with clear messages

   git commit -m "feat: add amazing feature"

6. Push and create a Pull Request

Contribution Ideas

• 🐛 Bug Fixes - Found a bug? We'd love a fix!
• ✨ New Features - Check issues for feature requests
• 📖 Documentation - Improve docs, add examples
• 🎨 UI/UX - Enhance the interface
• ⚡ Performance - Optimize inference, reduce memory
• 🧪 Tests - Add test coverage

Code Style

• Frontend: TypeScript, React hooks, CSS modules
• Backend: Python 3.10+, FastAPI, type hints
• Commits: Use Conventional Commits

Reporting Issues

When reporting bugs, please include:

• OS and version
• Node.js and Python versions
• Steps to reproduce
• Error messages/logs

---

📜 License

MIT License - see LICENSE for details.

---

🙏 Acknowledgments

• llama.cpp - Fast LLM inference
• Hugging Face - Model hosting

---

Built with ❤️ for privacy