The AI Bootcamp for L&D

Map every task in a phase of your workflow and tag each one: take it off me, make it faster, help me do it better, make it possible

Score tasks on frequency, pain and who'd notice — so you build where the payoff is, not where the hype is

Surface the ghosts: the tasks that should happen but never do because they're too big or complex

Check a task is buildable with three questions, then write a hypothesis with a countable pass mark in it — before you build

Write a product spec using my Four Rs framework, to make AI run to your method & standard, not its own

Build four tools live in class, one per phase of L&D: analysis, design, development, evaluation

Test each tool against real cases and the check the task already has — the template, the SME, the learner

Read the results against the pass mark you set — and find where the tool skipped, invented or averaged

Learn how to scale succesful tools using Projects, Skills, Agents or Apps

Spot the tasks & decisions you should never hand to AI — modality, scope, whether it should be a course at all

Choose the form — prompt, Project, Skill, Agent or App — and where the human check sits at each step up

Chain tools into a system and place a check at every join, for quality and accountability

Build learner-facing tools — a practice coach, a tutor, feedback bot etc — using my Four Rs framework to optimise reliability & impact

Decide where AI sits in the learning experience: what it can hold learners to, and what a human still has to see

Test a learner-facing tool with synthetic AI learners before it goes anywhere near a real learner

Ground every claim you make in the research — the failure modes, the evidence for each fix, and what's still unproven

Show your team a working tool, its test results and its pass rate, not a slide about potential

Take back a method anyone can follow — map, tag, validate, hypothesise, specify, test, scale