— Thinking —
Notes from the operators' floor.
Long-form essays on what we've learned. Some about AI. Some about businesses. All true. Written by Max, not by an agency content team.
Get on the calendar →or — read the manifesto below
— Nine essays · Nine published —
Why most AI consulting fails inside real businesses.
Most AI consultants have never been inside a real business. They've been inside AI. That, in one line, is why their work breaks the day it meets your Tuesday morning…
Read the manifesto →
01
★ Featured
Make it rule-based, and call the LLM as rarely as possible.
The most expensive mistake in an AI project is using the LLM for everything. The architecture that actually works is the opposite instinct…
Read the essay →
02
Technical
Most companies that say they are AI just wrap ChatGPT, and you can do the same thing yourself.
Very many companies that say we are AI just build a wrapper around ChatGPT and call it AI. There is no magic, and that changes how you should buy…
Read the essay →
03
Buyer's guide
Why you want a model-agnostic gateway, not one provider's chat.
Tie yourself to one provider and you get one provider's models. The ability to route any task to any model is the lever on both capability and cost…
Read the essay →
04
Technical
Staff augmentation vs traditional hiring, answered by someone who runs both.
I run both sides of this comparison. I hire engineers into my own company, and I place engineers into other people's teams, so every week I watch the same decision made from both ends…
Read the essay →
05
Buyer's guide
What a Magento operator learned that AI consultants get wrong.
Fourteen years of pricing rules and B2B tier logic and warehouse routing teaches you what AI is actually for. Most of our competition didn't go to this school…
Read the essay →
06
Featured
Outstaffing is not outsourcing, and the difference is the ceiling.
People use outstaffing and outsourcing interchangeably, and they are not the same business. They are not even similar businesses…
Read the essay →
07
Buyer's guide
Productionising RAG, honestly.
An eval discipline that killed BM25+RRF and parked a Cohere reranker. What survived, what didn't, and the unglamorous engineering between a demo and a product…
Read the essay →
08
Technical
A star on the CV who couldn't string two words together.
Complete fiasco yesterday. Great candidate on paper: every technology the client wanted, even a bit of the rare commercial AI experience they were hoping for…
Read the essay →
09
Staffing