Una potente IA. Direttamente sul tuo PC.
La tua prossima grande idea non ha bisogno di un data center. Esegui applicazioni di IA localmente sul computer che già possiedi. Personale, accessibile e alimentato da te.
What I'm building
I build the tools to work with AI. Today the focus is single hardware — one laptop, yours — but they're designed to scale easily to team-shared hardware and enterprise environments. One throughline: privacy as the first professional requirement, not an afterthought.
Single device
→
Team-shared
→
Enterprise
Inference Runtime
the layer below the model
A Metal-native LLM inference engine for Apple Silicon, written in Rust, talking directly to the GPU.
Agent Infrastructure
the layer around the agent
Memory, context, decision and typed Git that govern the throughput and reliability of the agents writing the code.
I measure, I don't guess. I certify, I don't trust.
Privacy-first AI
Work with AI without giving yourself away.
Your data stays yours. Every action auditable. Compliant by construction — local-first, private, under your control. You are the most important thing here, not the model.
The tools are open and local-first — try them, break them, tell me what's missing. Feedback and collaboration are welcome on GitHub.
Need to run part of your business with AI, privately and governed? The same foundation can grow into a tailored, private deployment for your organization. Tell me what you need — we build it together.
Build with me →