01
What to build
Which customer or team behaviour must change?
RuleWrite a hypothesis that can fail within 14 days. “Use AI” is not a hypothesis.
OutputOne sentence: We believe [change] will produce [outcome] for [person].
Founders · Bootcamps · University programmes
Turn an AI idea into a decision you can validate.
Resources for founders deciding what to build, automate and measure before an experiment becomes operating debt.
Designed so every session ends with
01
A falsifiable hypothesis
02
A 14-day experiment
03
Metrics that change decisions
Operating kit · 01
Three decisions before opening another tool
Use this as a one-page canvas. If an answer does not fit in one sentence, it is not ready to become a system.
02
What to automate
Is the task repeatable, reversible and observable?
RuleAutomate volume and repetition. Keep human judgement where errors are hard to detect or reverse.
OutputOne boundary: the system proposes, executes, or only prepares evidence.
03
What to measure
Which evidence would make the team stop, iterate or scale?
RuleMeasure cycle time, error/rollback, cost per useful outcome and validated learning.
OutputOne threshold per metric and its linked decision, agreed before starting.
Formatted for printing or saving as a PDF from the browser.
Protocol · 02
From classroom to a real test in 14 days
Days 1–2 · Frame
Define the user, friction, hypothesis, owner and one reversible decision.
Days 3–10 · Run
Use a small sample, visible traces and a manual fallback. Do not scale yet.
Days 11–14 · Decide
Compare with the baseline. Kill, iterate or scale, then document why.
For participants
You do not need to book consulting to get started.
Open the kit, run the protocol and return with evidence. When the bottleneck affects the whole system, the diagnostic helps select the right intervention.
Find the right intervention →For universities and programmes
Mentoring that leaves operating judgement behind.
Sessions, office hours, bootcamps and complementary modules that turn ideas into verifiable decisions without replacing academic teaching or the faculty team.
- → Startup mentoring and model reviews
- → AI Fluency and decision quality
- → Experiment and operating-model design
- → In-person, hybrid and online formats
The Operating Manual
Keep building after the bootcamp
The Operating Manual brings together signals, frameworks and decisions on AI, systems and growth. No noise.
Stay in the loop
Systems, strategy and signal. No noise.
No spam. Unsubscribe any time.
Bring a decision, not a deck.
We can review the experiment, architecture or operating model when there is a real decision to unlock.