Mechanical engineer (ETH Zürich · Harvard Biorobotics) building ML & edge systems —
and writing about how to think clearly with them.
I build systems — mechanical, digital, and mental — that make good work easier: machine-learning models that run on constrained hardware, a self-hosted homelab that just works, and a knowledge base operated hand-in-hand with AI. I trained as an engineer at ETH Zürich and did research at the Harvard Biorobotics Lab; today I care most about the seam where engineering, ML, and clear thinking meet.
If any of that is your world too, you're in the right place.
- layover — pulls flight confirmations out of your inbox and turns them into a clean, structured travel log. Deterministic parsing + LLM assist.
- An AI-operated knowledge system — a personal vault run as a partnership with AI, using a small "core-four" method (identity, current state, maps, capabilities). Writeup + open template coming.
- Homelab, as code — a documented, reproducible self-hosted stack (Docker · Proxmox · Tailscale).
- Edge ML, made approachable — a practical guide to deploying models on Google Coral hardware.
Building a second brain you actually use · self-hosting without the headaches · working with AI instead of around it. New pieces land at constantingemmingen.com — come read, and tell me what resonates.
Building at the intersection of engineering, ML, and clear thinking.
Have a problem worth solving, or just want to compare notes? Get in touch.