Builder @ Toceta — ex-C3 AI / Oracle. My whole life runs on a self-built AI OS: it triages my email, tracks my finances, preps my meetings, and holds a long memory of all of it — then grades my discipline in public, right here. Proof over vibes; bias for shipping.
This site is operated by my AI life-OS. Every number below is computed from a tamper-evident ledger, not typed by hand — the score can’t be faked, and the misses stay on the record.
round(100 × priorities-hit% × coverage%) how I compute this → consecutive sealed days where done ≥ 1 and missed_major = 0 · counts only days sealed into the ledger done ÷ deadlines set, trailing 30 days days logged ÷ available days · system outages excluded today − launch date Building AI-agent products and hardware — including the AI life-OS that runs this site. Shipping in public, and holding myself to it in the open.
- TocetaBuilder active — you are hereApr 2026 — present
Building AI-agent products and hardware — including the AI life-OS that runs this site. Shipping in public, and holding myself to it in the open.
AI agentsLLMsHardwareBuilding in public - what’s next
No priorities on record yet.
From the log
More in the log →How I grade my discipline
A number that grades my discipline is worthless if I can quietly fudge it. So the formula is public, it multiplies two things I can't both fake, and you can recompute it yourself from the same data the board shows.
type a command — e.g. score, explain 2026-08-08, help. Plain lines are facts; dimmed ~ lines are the agent's hedged guesses.
Or open a channel — code, the feed, a message, or the full dossier.
Follow the build