log
Writing
Engineering notes, architecture decisions, and post-mortems.
-
I tried to find a trading edge — and failed seven times. That is the point.
A pre-registered, bias-controlled research program found no robust edge across seven strategy families. Why a clean, well-documented negative is a success of method.
-
From storing logs to real observability
How instrumenting a critical algorithmic trading system turns log soup into questions you can actually answer in production.