Learn wide. Build sharp. Ship anyway.

Ten years in financial services data teaches you one thing early: the interesting problems never respect your job description. The pricing question turns out to be a data-quality question. The data-quality question turns out to be an incentives question. Learning wide is not a luxury — it is the only honest response to how problems actually arrive.

But breadth without an edge produces slideware. At some point you have to build the thing: the model, the tool, the pipeline that either works or doesn't. Building sharp means picking the smallest version that can be wrong in an informative way, and finishing it.

Why write here

This site is the notebook for that loop. Essays when something generalises, notes when it doesn't, build logs when the lesson only makes sense with the scars attached. Some of what I build lives here too — the public pieces on the apps page, the rest behind a door with one name on it.

Ship anyway

The last word of the hero line is the operative one. Nothing here waits for perfect. If it holds up, it ships; if it breaks, the breakage gets written up. That is the whole method.