WRITING / FIELD NOTES — EXHIBIT D

Field notes & writing

Field notes with plain-language summaries — and one bucket-list build, documented as a design record while it happens.

LOG 01

The ten-hour judging job we taught an AI to do fairly

Managers were spending ten-plus hours a day grading field work. Now an AI rates ~10,000 tasks a night, per dimension with the reason attached — proven against 1,000+ human judgments before it was allowed to, and kept honest by an appeal loop.

SHIPPED SYSTEM · WIFFY · ~10K NIGHTLY · 90%+ ACCURACY BAR
LOG 02

Designing multi-tenancy when your tenants are 100x apart in data size

How to let many companies safely share one software platform — each fully walled off, each fast, and none paying for their neighbor's size. With the audit dividend: separation you can demonstrate, not just promise.

SHIPPED SYSTEM · WIFFY · DB PER TENANT · BLUE-GREEN · AUDIT-READY
THE BUCKET-LIST BUILD — KEPT DELIBERATELY SEPARATE

GameMind AI: an AI that watches your gameplay and coaches you like a pro

The moonshot on my list, documented as a design record while it's being built — with the north-star metric and the kill criteria decided in advance, so the data outranks my attachment.

DESIGN RECORD · IN BUILD · NORTH STAR: REPEAT SESSIONS WITHIN 7 DAYSREAD THE DESIGN RECORD →
ALSO IN THE FILE — THE DESIGN-DECISION FAQ

Every box in my systems has a why behind it. Ten decisions written down — the constraint, the options rejected, the cost accepted.

OPEN THE FULL FAQ →