Private AI Infrastructure in One Command
Deploy a complete, secure, local AI environment on your own hardware. No cloud dependencies, no API costs, full data privacy.
curl -fsSL https://raw.githubusercontent.com/Rinkatecam/AI.Stack/main/install_aistack.sh | bashAI.STACK is a comprehensive installer that sets up a complete private AI infrastructure including:
- OpenWebUI - ChatGPT-like interface for your local AI
- Ollama - Run LLMs locally (Llama, Mistral, Qwen, etc.)
- Qdrant - Vector database for RAG and semantic search
- 12 AI Tools - Files, SQL, Web Search, Chemistry, Math, and more
- Infrastructure - Portainer, n8n automation, Paperless-ngx (optional)
All running on YOUR hardware, with YOUR data staying private.
# That's it. One command.
curl -fsSL https://raw.githubusercontent.com/Rinkatecam/AI.Stack/main/install_aistack.sh | bash- Automatically detects NVIDIA/AMD GPUs
- Installs appropriate drivers
- Selects optimal AI models for your hardware
- Works on CPU-only systems too
| Mode | Best For |
|---|---|
| Quick Install | First-time users, home labs |
| Custom Install | Specific requirements, experienced users |
| Secure Install | Production, enterprise, internet-facing |
- Three security levels (Basic, Standard, Hardened)
- Randomized internal ports
- Auto-generated secure credentials
- Firewall auto-configuration
- Localhost binding options
| Tool | Description |
|---|---|
| Files & Documents | Search, OCR, RAG with vector embeddings |
| SQL Database | Natural language to SQL queries |
| Web Search | DuckDuckGo, Wikipedia, Weather |
| Scientific Calculator | Math, unit conversions, equations |
| Chemistry | PubChem compound lookup |
| Visualization | Charts and graphs with matplotlib |
| Shell Execute | Safe command execution |
| Multi-Model Agents | Delegate tasks to specialized models |
| Document Templates | DOCX placeholder replacement |
| Code Analysis | Syntax validation, security scanning |
| Regulatory Lookup | FDA, ISO, MDR standards reference |
| Knowledge Base | Team experience database |
Pre-configured tool sets for different teams:
- R&D - Files, Chemistry, Math, Code, Visualize
- Regulatory Affairs - Files, Regulatory, Templates, WebSearch
- Quality - Files, SQL, Templates, KnowledgeBase
- HR - Files, Templates, KnowledgeBase
- IT - All tools
| Component | Minimum | Recommended |
|---|---|---|
| OS | Ubuntu 22.04+ | Ubuntu 24.04 LTS |
| RAM | 8 GB | 16+ GB |
| Storage | 30 GB | 100+ GB |
| GPU | None (CPU works) | NVIDIA 8GB+ VRAM |
- NVIDIA: GTX 1000+, RTX 2000/3000/4000 series
- AMD: ROCm-compatible (experimental)
- CPU: Works without GPU (slower)
curl -fsSL https://raw.githubusercontent.com/Rinkatecam/AI.Stack/main/install_aistack.sh | bashOpen your browser: http://localhost:3000
First user becomes admin. Create your account and start chatting!
# Check if your system is ready
curl -fsSL https://raw.githubusercontent.com/Rinkatecam/AI.Stack/main/install_aistack.sh -o install_aistack.sh
bash install_aistack.sh --check./install_aistack.sh # Normal installation
./install_aistack.sh --check # System compatibility check
./install_aistack.sh --resume # Resume interrupted installation
./install_aistack.sh --clean # Clean up failed installation
./install_aistack.sh -v # Verbose output
./install_aistack.sh -q # Quiet mode
./install_aistack.sh --help # Show all optionsAfter installation, these scripts are available in ~/ai-stack/:
| Script | Description |
|---|---|
status.sh |
Check service status |
restart.sh |
Restart all services |
backup.sh |
Backup with verification |
update.sh |
Check for updates |
pull-model.sh |
Add new AI models |
uninstall.sh |
Remove AI.STACK |
# Create verified backup
~/ai-stack/backup.sh
# List backups
~/ai-stack/backup.sh --list
# Verify backup integrity
~/ai-stack/backup.sh --verify /path/to/backup.tgz
# Restore from backup
~/ai-stack/backup.sh --restore /path/to/backup.tgzAll configuration is stored in ~/ai-stack/:
| File | Description |
|---|---|
docker-compose.yml |
Container orchestration |
ports.conf |
Port configuration |
secrets.conf |
API keys & passwords (delete after saving!) |
.install-config |
Full installation config |
After installation, save credentials from secrets.conf to your password manager, then delete the file:
cat ~/ai-stack/secrets.conf # View credentials
rm ~/ai-stack/secrets.conf # Delete after saving┌─────────────────────────────────────────────────────────────┐
│ User Browser │
└─────────────────────────┬───────────────────────────────────┘
│ :3000
┌─────────────────────────▼───────────────────────────────────┐
│ OpenWebUI │
│ (Chat Interface) │
└──────────┬─────────────────────────────────┬────────────────┘
│ │
┌──────▼──────┐ ┌───────▼───────┐
│ Ollama │ │ Qdrant │
│ (LLMs) │ │ (Vectors) │
└─────────────┘ └───────────────┘
│
┌──────▼──────┐
│ NVIDIA GPU │ (optional)
└─────────────┘
~/ai-stack/update.sh~/ai-stack/update-models.sh~/ai-stack/pull-model.sh llama3.2:latest~/ai-stack/status.shcd ~/ai-stack && docker compose logs -f~/ai-stack/restart.sh# Check NVIDIA driver
nvidia-smi
# Reinstall container toolkit
sudo apt install -y nvidia-container-toolkit
sudo systemctl restart dockerThe installer automatically finds free ports. If issues persist:
~/ai-stack/show-ports.sh~/ai-stack/uninstall.shOptions:
- Containers only - Keep data and config
- Containers + Config - Keep user data
- Everything except models - Fresh start, keep downloads
- Complete removal - Remove everything
Contributions are welcome! Please:
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
This project is licensed under the MIT License - see the LICENSE file for details.
- Ollama - Local LLM runtime
- OpenWebUI - Beautiful chat interface
- Qdrant - Vector database
- All the amazing open-source AI models
- Issues: GitHub Issues
- Discussions: GitHub Discussions
Built with care by Rinkatecam & Atlas
Your AI, Your Hardware, Your Data