Local AI
AI that runs on my own hardware, from language models to autonomous agents.
Overview
Building and experimenting with AI that runs on my own hardware. This lab explores how capable local language models, autonomous agents, voice assistants, and self-hosted infrastructure can become.
Problem / idea
Cloud services make powerful AI accessible, but they also make the underlying system distant. Local AI is a study in capability, privacy, control, and the limits of the hardware in front of me.
What I built
A growing local stack of language models, autonomous agent workflows, voice experiments, and self-hosted services with explicit logs and small, testable scopes.
Challenges
The system has to be useful without hiding its constraints. Latency, memory, power, and failure modes are part of the interface, not details to smooth over.
Stack
- AI
- Agents
- Automation
Documentation in progress. Process notes and build logs land here as the work moves.