PEO · Internal AI Enablement

The AI Teaching Method

How we teach our teams to use AI on real work — วิธีที่เราสอนทีมงานให้ใช้ AI กับงานจริง

Staff AI-training session in a Thai office, laptops open with screens glowing
Same foundation for everyone. Different use-cases per role.

พื้นฐานเดียวกันทุกคน · ตัวอย่างงานต่างกันตามบทบาท — Core (~70%) is identical for all; the use-case layer (~30%) is swapped per role.

The core curriculum

  1. AI is a brilliant intern — fast and tireless, but it sometimes guesses wrong (check every time) and it needs clear instructions.
  2. Three rules: always check · give clear instructions (context + task + format) · keep company data in company tools (company Workspace AI only — never personal AI tools for company data).
  3. Prompt formula: context + task + format = result.
  4. Data safety: safe = work emails, internal drafts, documents · never = payroll/personal data, customer financials.
  5. Email & document discipline: let AI draft → you read, check, correct → then send. Summarize instead of forwarding screenshots. Translate on demand.

Why we teach this way

Every participant can make their work visible, reusable, reviewable and reportable.

The goal is a durable change in how work gets done, not just awareness of AI.

How the program builds

  1. Orientation, safety, and a first real task
  2. Email & document discipline
  3. Gemini Gems — durable role memory for a recurring task
  4. Gem refinement + a checking checklist
  5. Weekly personal work report
  6. Prompt quality + spotting hallucination
  7. Real-work workshop (prompt / output / human-final)
  8. Source-grounded summaries — only where useful and safe
  9. Team weekly report (management view)
  10. SOP / knowledge extraction from repeated reports
  11. Dashboard rhythm
  12. Final cumulative report — tied to the person’s job

Gemini Gems

A “Gem” is a saved AI helper with a fixed role, context, task, format, limits, and a self-check rule — an AI that already knows your recurring job, instead of one you re-explain every time. A good Gem always ends by flagging what it was unsure about. The human still checks.

The weekly report

It escalates. Start with three simple questions — what did AI help with? did it work? what will you try next? — and grow it week by week until it ties to the person’s actual job. This makes each person’s AI work visible and reviewable.

How we measure maturity

An operating scale, not school grades:

0–1no evidence → awareness
2–3real task → shows prompt & checks
4human-corrected final
5builds a reusable Gem
6weekly cumulative report
7contributes to team SOP

The one requirement

Apply it to real work.

A summary about AI is not the goal. The goal is AI applied to a real task, checked by a human, and made reusable.