Field notes from
the work itself.
Practical writing on data strategy, analytics engineering, and applied machine learning — drawn from real engagements, written for the people who have to act on the results.
The question comes first: why most analytics projects fail before the first query
The hardest part of analytics is rarely the maths. It's deciding what question is actually worth answering — and most teams skip it. Here's how we frame an engagement before a single query is written.
The dashboard nobody opens: designing reporting around decisions
A dashboard that nobody opens is a cost, not an asset. The problem is almost never the data — it's that the thing was designed around what's available rather than what someone has to decide.
Build vs. buy: a framework for analytics infrastructure decisions
Build-versus-buy is rarely a technology question. It's a question about where your differentiation actually lives — and where you're quietly signing up for maintenance you'll regret.
Production is where the real work starts
The notebook with the impressive accuracy score is the start of the work, not the end. Most of the value — and most of the risk — lives in everything that happens after a model ships.