How we think about the work.
Practical insights on data, software, geospatial systems, and AI — written from the work, not for a content calendar.
Your mining data room is a shared folder — here's what that costs
Most junior explorers run their "data room" out of a shared drive. It works until it doesn't — and the failure always shows up at the worst possible moment.
Arguments, decisions, and working patterns from the systems behind the work.
A fixed price is a forcing function on us, not you
Why we quote fixed prices for defined work: not as a marketing gesture, but because it forces us to understand the problem before we charge for solving it.
The pyramid is optional
The classic consulting pyramid existed for one reason: leverage. Now that production-layer work is automatable, we run four senior people, an agentic production system, and a hard rule about who signs off.
When to build a QGIS plugin instead of a web app
A plugin lives inside the tool your analysts already use. Here is an honest test for when that beats a web app, a platform, or a spreadsheet.
The busywork tax — the hours small teams lose to manual work
Re-keying data, copy-pasting into reports, chasing files, checking a site for changes. Most of it is automatable now — but not all of it is worth automating.
Why your website is invisible to AI search (and what to do about it)
Most business sites are client-rendered single-page apps. Google tolerates that; ChatGPT, Claude, and Perplexity do not. Here is why, and the fix.
Small tools, big leverage
Not every problem needs a platform. Often the highest-leverage software is small and shaped to one job — matched to where the work actually happens.
ArcGIS config vs. a real geospatial platform — what teams get wrong
"We already have GIS" and "we have a mapping system" sound the same and mean completely different things. Confusing them is expensive.
If you can't measure it, you bought a vibe
AI systems ship on the strength of a demo, then quietly fail on the cases that matter. The evaluation harness isn't overhead — it's the product decision.
Why two reports from the same company give different numbers
When finance says one thing and operations says another, the problem is rarely the spreadsheet. It's that nobody agreed where the truth lives.
How to structure exploration data so it doesn't rot
Claims, assays, and drill results don't decay because the data is wrong. They decay because there's no model holding them together. Here's the model.
Bring us the version that still feels messy.
One senior conversation is enough to establish whether there is a useful place to start.
Start a fit conversation →