PROJECTS / 03
ACTIVE EXPLORATION
Practical local AI, documented honestly.
Hands-on exploration of AI systems for the home lab: practical infrastructure, repeatable experiments, and systems that do real work — tested on real hardware with real software.
Conceptual artwork — a compact compute rack. Not a BigSteve Labs facility.
01 · WHAT IT COVERS
The full stack of a working home lab.
The Home AI Lab is where ideas become working systems. Projects span the practical layers of local AI:
- Local large language models — running and serving LLMs on-site
- Ollama and Open WebUI deployments
- ZimaOS home servers
- Windows AI assistants
- AI-powered business automation
- Home networking and remote access
- Docker and containerized services
- MCP servers and agent frameworks
- Open-source AI tools
- Voice assistants and speech systems
- Hardware testing and optimization
02 · BUILT FOR PRACTICAL AI
Systems that perform useful work.
Many AI demonstrations stop after generating text. This lab focuses on systems that actually perform work: managing business workflows, organizing information, automating repetitive tasks, monitoring servers, testing new frameworks, and running private AI infrastructure without depending entirely on cloud services.
THE STANDARD
Every project is documented with real successes, real mistakes, and real solutions — the goal is repeatable processes others can follow, not showcase demos.
03 · ALWAYS LEARNING
Failures are documented too.
Technology changes every week, and the Home AI Lab is never considered finished. New hardware, new models, new software, and new ideas are constantly tested. Some experiments succeed. Others fail spectacularly. Both are valuable — every experiment teaches something that can be applied to the next project. This page is a living journal of that process.
The assistant that runs on this infrastructure.
→
The lab notebook’s current directions.
→