AI workshop intake template for teams worksheet

how to use

Run it once, then improve it.

Copy the worksheet into Notion, Google Docs, Linear, or a plain text file. Fill it with sanitized examples only. After the first run, update the failure notes before sharing it with anyone else.

Human gate: An organizer must read every intake reply before clustering. LLM may suggest themes but cannot decide who attends or what is exposed.

Privacy: Do not paste real customer messages, health, or financial data into the LLM. Use sanitized examples or schemas only.

# AI workshop intake template for teams worksheet

Source workflow: https://dabblewith.ai/workflows/ai-workshop-intake-template-for-teams/

## 1. The real job
- Who is this for?
- What recurring task are we improving?
- What does a good output look like?

## 2. Safe input
- What sample input can we use?
- What must be redacted first?
- What data must never enter an external AI tool?

## 3. Workflow draft
Problem: You're running an internal AI workshop and don't know what each attendee actually wants to build, so the session ends up generic.

Outcome: Each attendee shows up with a specific workflow problem, sample inputs, and a publish-or-not decision before the workshop starts.

Tools today:
- WhatsApp
- Email
- Spreadsheet
- Any LLM with a draft-only review pass

Steps:
1. Define the workshop's single outcome — a published workflow, a working draft, or a reviewed prompt.
2. Send an intake message asking: role, one workflow they want, sample input they can share safely, what they'd never automate.
3. Use an LLM in draft-only mode to cluster responses into 3-5 themes. Read every response yourself.
4. Pre-write a starter workflow for each cluster so no one starts at a blank screen.
5. Confirm with each attendee 24 hours before the session that the starter fits their problem.

## 4. Human review gate
An organizer must read every intake reply before clustering. LLM may suggest themes but cannot decide who attends or what is exposed.

Reviewer:
Approval rule:
Escalation rule:

## 5. Privacy / never automate
Do not paste real customer messages, health, or financial data into the LLM. Use sanitized examples or schemas only.

Never automate:
- 
- 
- 

## 6. First run notes
What worked:
What failed:
What changed after review:

## 7. Publish / share decision
Can this be published? yes / no
What must be removed before publishing?
Attribution / backlink preference: