A four-part programme that takes a team from first prompts to a working AI operating system.
This is a structured curriculum in how to work with AI at work, and the same method we use when we roll it out inside our own companies. It is general and educational, with no vendor pitch. You start by treating AI as a capable intern whose work you check every time. You finish running an operating system that makes work visible, reusable, and reportable, with Claude as a coworker and Claude Code as an operator. Throughout, the discipline is step four of the P1 method — keep ownership and accountability clear: the system carries the work, but the decisions stay with the people who own them.
The programme is sequenced: the foundation → reusable helpers → a coworker you delegate to → an operator that works on your files → your own AI operating system. Each part builds on the one before it. Work through it in order, or start at the stage that matches your team.
01
Treating AI as an intern whose work you check. The three rules, the prompt formula, data safety, reusable helpers, and the weekly report.
→
02
Delegating real work to Claude — Projects for saved context, Artifacts for live deliverables, and the checks that keep the output reliable.
→
03
AI that works directly on your files and systems, with a person approving each action and the guardrails that requires.
→
04
A source of truth, workflows built around it, and a review layer that keeps it accurate — where the coaching connects to our advisory practice.
→This is the AI training curriculum P1 Thailand built and teaches its own team with, now offered to other organisations. All three tracks share the same length, pace, and safety principle; each one trains a different capability. Pick the track that matches where your team is today, or run all three in sequence.
Working fluency with AI: choosing the right method, writing prompts that hold up to checking, verifying sources, and building helpers you reuse. The entry point for a team starting from zero.
12 weeks → TRACK 02Moving from instructing AI to thinking with it: framing problems, separating fact from assumption, stress-testing ideas, and building a reasoned, evidenced recommendation.
12 weeks → TRACK 03A decision map for AI: capability categories, model limits, data governance, and evaluating tools on evidence instead of chasing names.
12 weeks →The same shape, three times. Every track runs 12 weeks at 90 minutes a week, for groups of 6–20, in the same demo → practice → applied-to-real-work → reflect rhythm. And every track holds the same line: no customer, personal, or financial data goes into an unapproved tool — training or anonymised examples only, and the person using AI always owns the decision and the result.
Alongside the three tracks above, the maturity ladder shows where each person is today, from no experience to setting team-wide standards — the same measure across all three. Full detail lives on the foundation page.
Three supporting pages sit alongside the four parts: a plain primer on how AI actually works, the full programme structure, and how it applies role by role.
A plain, non-technical primer — models, next-word prediction, and why AI makes confident mistakes.
The whole programme at a glance — the design rule, the three twelve-week tracks, the weekly report, the tooling path.
The same foundation, then the use cases that fit back-office, sales, finance, operations, and site teams.
The curriculum is how our own family business group operates day to day — the same foundation, discipline, and operating system. We run it in our companies first, then teach it.
One shared foundation for the whole team, applied to each person's real work.
Request a consultation