HOW WE WORK
Three stages of AI work, in order. Free, then paid, then production.
Every engagement runs the same shape: a free diagnostic, a paid map, an optional fixed-fee build. We measure before we build, and we only build what the measurement justifies.
01 · DIAGNOSE
The Readiness Score
A 20-minute conversational diagnostic that probes four pillars — data, process, people, strategy — and returns a calibrated 0-100 score plus a written report. You keep the report whether you hire us or not.
FORMATCONVERSATIONAL, ~10 MIN
OUTPUTSCORE + WRITTEN REPORT
PRICEFREE
WHENWEEK 0
The diagnostic is calibrated against the businesses we’ve actually scoped. A low score isn’t a rejection — it’s a list of what to fix before AI is worth spending on. Sometimes the answer is “clean up your CRM and come back in six months,” and we’d rather tell you that for free than sell you a build that fails.
02 · PRESCRIBE
The Application Map
Two weeks of structured discovery: workshops with your team, a review of your data and tooling, and a written 12-20 page Application Map that ranks every viable AI workflow by payback, complexity, and failure mode.
FORMATWORKSHOPS + WRITTEN REPORT
OUTPUTRANKED ROADMAP, 12-20 PAGES
FROM$8,000
WHENWEEKS 1-2
The map is drawn from our pattern catalog of 42 workflows we ship, plus the patterns we won’t build (autonomous sales agents, customer-facing regulated advice, custom fine-tuning at SMB scale, “AI to replace your team”). For every recommendation we publish the failure modes we’ve seen elsewhere — not as marketing, but because writing the failure mode in is how we keep ourselves honest at scoping time.
03 · BUILD
The Implementation
Production builds of the workflows the map prescribed. Fixed-fee, instrumented from day one, with a written kill-criterion: the specific metric that would tell us to shut a workflow down.
FORMATFIXED-FEE PROJECT
OUTPUTPRODUCTION WORKFLOWS + INSTRUMENTATION
FROM$24,000
WHENWEEKS 3-14
Every Build engagement we sign includes a written kill-criterion. An example: “If deflection rate stays below 30% after 60 days in production, or CSAT on AI-handled tickets drops more than 0.4 points from baseline, we’ll recommend killing this workflow rather than continuing to optimize.” Most consultancies prefer to keep options open — we don’t. We’d rather kill a $24,000 workflow at month three than keep billing to maintain it for two years. That decision is what we sell.