A modular, reproducible and production-oriented local AI laboratory built with Docker.
The goal of this project is to design and maintain a local environment for developing, deploying and experimenting with AI services, coding agents and large language models using infrastructure engineering best practices.
Build a reusable AI laboratory that can evolve from a single coding agent into a complete AI platform.
The laboratory is designed around the following principles:
- Containerization
- Infrastructure as Code
- Reproducibility
- Documentation
- Version Control
- Incremental Architecture
- Learn Docker through real-world deployments.
- Build a reproducible local AI environment.
- Deploy AI coding agents.
- Run local Large Language Models.
- Experiment with AI frameworks.
- Apply software architecture principles.
- Document every technical decision.
Current Phase:
Phase 0 — Environment Audit
Status:
🟡 In Progress
- Claw Code
- Ollama
- Open WebUI
- PostgreSQL
- Redis
- Custom AI Agents
ai-local-lab/
│
├── docs/
├── adr/
├── compose/
├── services/
├── stacks/
├── scripts/
├── volumes/
└── backups/
Project documentation can be found inside the /docs directory.
Architecture decisions are documented under /adr.
- Phase 0 — Environment Audit
- Phase 1 — Infrastructure Foundation
- Phase 2 — Docker Platform
- Phase 3 — Claw Code Deployment
- Phase 4 — AI Model Integration
- Phase 5 — Supporting Services
- Phase 6 — Custom AI Agents
MIT License