SD AI Training
All modules

Module 4 of 6 · 7 min

Hands-on walkthroughs

Last module: Five everyday wins, one safe loop: you give context, AI drafts, you review, then it goes out.

Time to actually do it

Reading about AI is like reading about swimming. Open your AI tool in another tab and do these three exercises with your own material. Budget ten minutes.

Exercise 1 — the email you've been putting off

Everyone has one: a delay to explain, a price increase, an estimate to decline. That’s your test subject.

  • Step 1. Tell the AI what’s going on, in plain unpolished words. Include the facts that matter.
  • Step 2. Ask for a draft: tone, length, what it must and must not say.
  • Step 3. Give feedback at least once — “shorter,” “less formal,” “don’t promise a date.”
Starter — replace the brackets with your real situation
Here's the situation: [what happened, in your own words — don't polish it].

Draft an email to [who]. It needs to [the one thing it must accomplish]. Keep it [short/friendly/firm], and do not [the thing you don't want promised or mentioned]. Sound like a real person.

What good looks like:a draft you’d send after changing a sentence or two. If it’s generic, your brief was generic — add the missing details and go again.

Exercise 2 — turn what's in your head into an SOP

Every business runs on know-how stuck in one person’s head. Writing it down never happens, because writing documentation is miserable. So don’t write. Talk, and let the AI structure it.

  • Step 1. Pick one process you know cold — closing up, onboarding a customer, handling a return.
  • Step 2. Explain it like you’re training someone. Use your phone’s dictation, or type fast and messy. Include the exceptions — “unless it’s a Friday delivery…” — that’s the real knowledge.
  • Step 3. Paste the ramble with this prompt:
Ramble → SOP
Below is me explaining, out loud and unorganized, how we do [process] at my business. Turn it into a clean step-by-step SOP a brand-new employee could follow:

- Numbered steps in order
- Call out the exceptions and edge cases as notes on the relevant step
- A short "common mistakes" section at the end
- Ask me about anything that's unclear or missing rather than guessing

[paste your ramble]

What good looks like:the AI asks two or three sharp questions. Answer them — that’s a week of “I’ll get to it” done in ten minutes.

Exercise 3 — make sense of some numbers

Grab something real and small: last quarter’s sales by month, hours by job. Copy it straight out of the spreadsheet — messy formatting is fine.

First-pass analysis
Here's [what this data is] from my [type of business].

1. Describe what you see in plain English.
2. What are the 2–3 things worth my attention?
3. What would you want to know next that this data doesn't show?

Cite the specific numbers behind each point.

[paste your data]

Knowledge check

Answer all three to complete the module. Wrong guesses just mean try again.

1. In the SOP exercise, why do you talk through the process out loud (voice or rambling text) instead of writing it directly?
2. When you paste sales numbers in and ask for insights, what should you do with a surprising claim like "Tuesday revenue dropped 40%"?
3. What's the common thread across all three walkthroughs?