Sai vs. Reality
Building in public. Failing with documentation.
I ship software into unrelated categories — pool practice, breathing, Bible reading, news — and Reality tells me which ones were wrong. Those stay published too, with the evidence.
Products
Every one of them, with what it actually is and where it actually stands.
-
Runout RankLive
A pool rating from ten generated table layouts instead of 200 league games.
-
Pool & Billiards Self-TrainerLive
Drills, skill exams and match tracking for practising pool on your own.
-
Snooker Self-TrainerLive
The same structure as the pool trainer, rebuilt for a twelve-foot table.
-
Bible ProjectLive
KJV and ESV offline, with an AI that answers when you have a question.
-
FlexiLive
Timed, illustrated stretching plans with voice guidance. Built for a habit, not a workout.
-
TacetLive
Breathing paced by vibration with the screen off, so nobody in the room notices.
-
Calmly NewsLive
Ten stories a day, stripped of drama adjectives. Then the feed locks.
- Sai StudioLive MVP builds for non-technical founders. The consulting arm that pays for the rest of this.
- ChatengageBuilding Send documents as one package with a bot that answers questions about them.
- Harness SurveyLive A survey that reads each answer and writes the follow-up question itself.
Killed
Products Reality closed. They keep their URL and gain a post-mortem: what I believed, what I built, what the evidence said, what I would do differently.
Why any of this
Build less. Learn faster. Kill sooner. Try again.
I'm not trying to become successful. I'm trying to become less stupid. Building a lot of small things across unrelated categories is the cheapest way I have found to be wrong quickly and in public. More about how this works →