Insights

Notes from the field.

Engineering, strategy, and operational lessons from the work we ship. Opinionated and specific. No "AI is transforming everything" intros.

Filter
Engineering · 2026-07-14 · 6 min

Context engineering is the new systems design.

Prompt tips are a distraction. The hard problem is deciding what the model gets to see, and that's an architecture problem, not a wording one.

Tactical · 2026-07-07 · 5 min

The token bill nobody forecasts.

The demo cost thirty cents. Production costs forty grand a month. The gap is made of calls nobody put in the spreadsheet, retries, agent loops, eval runs.

Product · 2026-06-30 · 5 min

Human-in-the-loop is a design primitive, not a fallback.

Most teams bolt a human onto the failure path. The teams whose AI gets trusted design the handoff first, as a first-class part of the system.

Industry · 2026-06-23 · 6 min

Shipping AI in regulated industries.

Healthcare, finance, and pharmacy don't reward the fastest model. They reward the system that can explain itself when the regulator asks.

Strategy · 2026-06-16 · 6 min

How to actually measure the ROI of an enterprise AI deployment.

"Productivity is up" is not a number a CFO can bank. How to measure AI ROI so it survives the budget review, and gets you the next round.

Engineering · 2026-04-26 · 6 min

Why your AI agent works in the demo and breaks in production.

Most agent demos run for 30 seconds in a sandbox. Production agents have to work for years. Four patterns that hold up.

Engineering · 2026-04-22 · 5 min

RAG is not a product. It's an architecture.

Most teams treat RAG as plumbing. It isn't. The failure modes live in places that never show up in a demo.

Engineering · 2026-04-15 · 5 min

The eval is the product.

Without evals you can't deploy with confidence, can't improve with discipline, can't catch the regression that's about to embarrass you.

Strategy · 2026-04-08 · 6 min

Build, buy, or wait: a framework for enterprise AI investment.

The interesting AI initiatives are never trivially "buy" or "build." They're the ones in between.

Strategy · 2026-04-01 · 6 min

The four ways enterprise AI engagements fail.

Wrong problem, wrong data, wrong adoption, wrong governance. Spotting them early is the difference between deployed and shelved.

Engineering · 2026-03-25 · 5 min

Frontier models vs. open-source: when each one wins.

Most production systems we ship use both. The question isn't "which one." It's "where does each one earn its place."

Strategy · 2026-03-18 · 5 min

What to ask when interviewing an AI consultancy.

Most consultancies sell strategy slides. Three questions separate builders from deck-makers.

Strategy · 2026-03-11 · 5 min

The hidden cost of generic AI SaaS.

You don't pay in license fees. You pay in workflow drift, and that bill compounds.

Tactical · 2026-03-04 · 5 min

How to run an AI proof-of-concept that doesn't waste a quarter.

A POC that can't ship is an expensive demo. Three rules to make yours actually inform the decision.

Product · 2026-02-25 · 5 min

Designing AI for people who hate it.

Frontline workers, clinicians, analysts, the people who actually use enterprise AI often start out skeptical.

Industry · 2026-02-18 · 6 min

Notes from a 200,000-employee frontline AI deployment.

Frontline AI at retail scale isn't a feature. It's a platform decision. Three lessons.

No posts in this category yet. Show all posts →

More posts publishing weekly. Subscribe to be notified, or check back.