Pattern Analysis
Periodic step-backs to look for patterns across months of work: what shifted, what stayed constant, and what that says about the direction.
Most recent entry
September 2026 — Third EntryStepping back across the last few months: a behavioural decision-intelligence platform taken to build-complete, a working paper published to SSRN, two markets job simulations passed, and a regulatory exam cleared. On the surface these are four unrelated things. Read together, they are not.
The recurring subject is the gap between how decisions are supposed to be made and how they actually get made. Kynexis measures that behaviour directly. The paper measures one structural version of it, cash-flow volatility interrupting the compounding that should otherwise happen. The simulations rehearse the decisions and the desk work is where real people make them. I did not set out to pick a theme. It is just what everything I chose to build turned out to be about.
What shifted since the last of these step-backs is the centre of gravity. The earlier stretch was mostly consuming: courses, certifications, reading. These months are mostly producing: a product, original research with a protocol so it can be reproduced, and a registration exam the industry actually requires. The credentials did not stop. They stopped being the point.
What stayed constant is that none of it is abstract. Real XBRL filings, real backtests, a live trade validated end to end, real clients. When I reach for a theory now, I only seem to trust it once it has touched something real. That has not changed, and I do not want it to.
Browse by date
Circled dates have an entry. Use ← → to scroll back through the months.