Practical decisions for AI and software
Technical guides to choose models, verify software and plan delivery.
Search every article
Search the full article text.
No posts match your search. Try another term or pick a topic.
Latest articles
New comparisons, practical guides and engineering lessons.
How can a water utility field-test a LoRaWAN meter reading and leak alarm?
A water meter can deliver a reading while its leak alarm is still untested. What to ask a supplier to show at two sites before placing a larger order.
Pivot the MVP or change the user flow? The Offlinery decision
When customers use an MVP differently than planned, what warrants a pivot? How Wavect and its client used that evidence to change one step at Offlinery.
Both software vendors say they’re done. What should you test before signing off?
A customer paid and the operator had no order. Follow one transaction across two systems and decide what evidence is worth asking for before release.
Does a fintech MVP need a market data API, or will Excel exports do?
Our first report helped asset managers judge the timing and price of a bond issue. Behind it sat a quarterly Excel export the startup already licensed.
Is AI Bringing Waterfall Back? How Much Upfront Planning Is Enough?
Faster implementation changes the case for planning. Use uncertainty, reversal cost and a stopping rule to decide what to settle before an agent builds.
Small model fine-tuning: Airbnb’s production lesson
A specialist classifier needs more than a cheap inference call. Evaluate training data, accepted errors, fallback, full costs and the path to production.
img2threejs: Turn an image into editable Three.js code
Start with the GitHub skill, inspect its TypeScript output and test the changes your browser prototype needs. Includes setup and a practical reconstruction prompt.
Organoid: AI material specification with evidence attached
Independent Organoid AI ideas for product answers and evidence packs. Hypothetical workflows with checks for carriers, mounting and document dates.
ING-B: AI for electrical engineering document review
Independent ING-B ideas for tender revisions, lighting evidence and handover checks. Hypothetical workflows, with no company AI implementation or results claimed.
Get the next field note
One concise email when we publish. No tracking pixels, and no inbox filler.
Facing one of these decisions right now? Talk it through with us →