Insights

Field notes from AI, geospatial, and engineering work

Practical write-ups from the teams actually delivering the work.

Jul 18, 20267 min read

The gap between a model that works and a model that ships

Most ML programs don't fail on accuracy. They fail on everything between the notebook and production - and that gap is an engineering problem with a known shape.

Jun 30, 20266 min read

Designing GIS field tools for zero connectivity

Offline-first isn't a checkbox - it changes your data model, your sync strategy, and your conflict resolution from day one.

$The first 90 days of a fractional CTO engagement - Fractional Leadership$

Jun 12, 20266 min read

The first 90 days of a fractional CTO engagement

What a fractional executive should actually deliver before day 30, 60, and 90 - and how to tell early if the engagement is working.

May 22, 20266 min read

Scoping an MVP that answers a question instead of hedging

Most MVPs are too broad to build quickly and too shallow to learn from. The fix is deciding what question you are actually asking.

May 2, 20266 min read

Data quality tests are more valuable than another dashboard

What changes operationally when pipelines fail loudly on bad data instead of passing it downstream to a chart nobody trusts.

Apr 14, 20267 min read

Measuring ROI on immersive training programs

The metrics that convince a budget committee immersive training is worth it - and the ones that don't.

Get insights in your inbox

One email a month - new engagement write-ups, no filler.