Build a Small-Business AI Operator

Step 1 — Choose one bottleneck

Pick a repeated task that consumes time and has a clear output. Do not start by asking AI to “run the business.”

Step 2 — Map the workflow

Write the trigger, inputs, steps, decision points, output and owner. You cannot automate a process you do not understand.

Step 3 — Split human and AI work

Use AI for drafting, classification, summarization and options. Keep judgment-heavy, sensitive and consequential decisions with a person.

Step 4 — Create reusable assets

Turn repeated work into templates, checklists, response libraries and structured briefs so quality does not depend on memory.

Step 5 — Add a review gate

Customer-facing claims, prices, financial actions, legal statements and promises should be reviewed before release.

Step 6 — Measure one outcome

Choose a practical metric: response time, hours saved, leads followed up, turnaround time, or error rate.

Step 7 — Improve before expanding

Run the workflow repeatedly. Fix failures and edge cases before adding more autonomy or another agent.

Copy the complete blueprint

Customize the bracketed details for your situation, then test it on a real but low-risk task before depending on it.

You are my small-business operating assistant. Help me turn one business objective into a practical weekly execution plan.

Start by asking for:
- the business and offer
- the customer
- the objective
- available time/budget
- current channels/assets
- important constraints

Then:
1. Identify the highest-value bottleneck.
2. Recommend the smallest useful workflow to improve it.
3. Break the workflow into human tasks, AI-assisted tasks, and tasks that can be automated safely.
4. Create the needed drafts/templates.
5. Define one measurable result for the week.
6. Flag legal, financial, privacy, customer-service, or brand decisions for human review.

Do not manufacture customer research, testimonials, results, prices, legal claims, or performance numbers.

The Human Edge Check