AIforBC Learning · Curriculum 1.0

AI for Teams
Facilitator Guide

A complete run-of-show for moving one team from curiosity to a controlled 30-day workflow pilot.

Open buyer readiness guide
Facilitated product

Adapt examples and pacing, but preserve the information boundary, independent verification, accountable human review, baseline, stop conditions, and dated decision.

Before delivery

Do not enter the room without a defined workflow owner.

  • Confirm the buyer, decision authority, participant list, delivery format, timing, accessibility, and accepted commercial scope.
  • Name an accountable workflow owner and an escalation owner for privacy, security, legal, employment, safety, or professional questions.
  • Confirm which AI tools and information classes are approved for demonstrations and practice.
  • Select one fictional demonstration and up to three sanitized candidate workflows.
  • Ask participants to record the current process, full-cycle time, review effort, errors, and common exceptions.
  • Test display, audio, captions, links, demonstration account, workbook printing, and a non-AI fallback.
  • Open the private cohort record and confirm that the evaluation link is active.

Opening script

Set an honest promise in the first five minutes.

Say

“Today is not about automating everything or trusting an AI answer. We will choose one bounded workflow, practise with approved information, verify the result, keep a human owner, and define evidence for a 30-day decision.”

Group agreements

  • No passwords, personal records, customer files, confidential documents, or unapproved information in demonstrations.
  • Participants may pause, observe, or use fictional information.
  • Confident language is not evidence.
  • The accountable person—not the AI system—owns every consequential decision.
  • Questions outside the workshop boundary are recorded and referred to the appropriate owner.

Module 01 · 25 minutes

Set the boundary

Observable outcome

The team agrees which information, tools, and decisions are approved, conditional, or prohibited.

Run-of-show

  1. Frame the operating outcome and human owner — 4 min
  2. Explain useful, unreliable, and consequential AI work — 5 min
  3. Sort sample tasks into green, yellow, and red zones — 9 min
  4. Draft the team boundary card — 5 min
  5. Confirm unresolved policy questions and owners — 2 min
Evidence to leave behind

A boundary card with approved tools, prohibited inputs, required review, and a named escalation owner.

Stop condition

Do not infer the organization’s legal, privacy, cybersecurity, employment, or professional rules. Record unresolved questions for the appropriate owner.

Module 02 · 35 minutes

Give useful instructions

Observable outcome

Participants turn a vague request into a bounded task brief with context, evidence, constraints, output, and review.

Run-of-show

  1. Show a vague request and name what is missing — 5 min
  2. Introduce the six-part task brief — 7 min
  3. Facilitator demonstration using fictional data — 7 min
  4. Pairs rewrite a team-relevant request — 10 min
  5. Compare outputs and revise the brief — 6 min
Evidence to leave behind

One reusable task brief that another participant can understand and review.

Stop condition

Use synthetic, public, or explicitly approved examples. Remove identifying customer, employee, financial, credential, and confidential information.

Module 03 · 35 minutes

Verify before relying

Observable outcome

Participants separate claims from evidence and apply a repeatable output-review checklist.

Run-of-show

  1. Present a confident but flawed answer — 5 min
  2. Identify claims, assumptions, and missing evidence — 8 min
  3. Open and compare original or authoritative sources — 8 min
  4. Record corrections and uncertainty — 8 min
  5. Assign the accountable reviewer — 6 min
Evidence to leave behind

A completed review checklist with sources, corrections, exceptions, and a human approval decision.

Stop condition

Do not ask the model to grade its own truthfulness. Important claims require independent evidence and a qualified reviewer.

Module 04 · 40 minutes

Choose the first workflow

Observable outcome

The team selects one frequent, bounded, reviewable workflow rather than a broad transformation project.

Run-of-show

  1. Generate three recurring candidate workflows — 7 min
  2. Define current friction and full-cycle baseline — 8 min
  3. Score frequency, value, feasibility, and consequence — 12 min
  4. Challenge hidden review and exception work — 7 min
  5. Select one workflow and one owner — 6 min
Evidence to leave behind

A prioritized workflow scorecard with an explicit no-go reason for alternatives.

Stop condition

Reject a first pilot that changes rights, safety, employment, payment, regulated advice, or external commitments without qualified review.

Module 05 · 35 minutes

Design the controlled pilot

Observable outcome

The team defines inputs, expected output, review, exceptions, baseline, measure, and stop conditions before changing work.

Run-of-show

  1. Map the current process and accountable owner — 7 min
  2. Map the assisted process and review cycle — 8 min
  3. Define approved inputs and expected output — 6 min
  4. Choose one outcome and one risk measure — 7 min
  5. Write exceptions and stop conditions — 7 min
Evidence to leave behind

A one-page pilot brief that can be approved or rejected without a sales narrative.

Stop condition

Do not use time saved as the only measure. Include correction time, review time, rework, exceptions, and downstream consequences.

Module 06 · 30 minutes

Operate for 30 days

Observable outcome

The team leaves with practice cadence, owners, evidence, and a dated expand-revise-stop decision.

Run-of-show

  1. Assign workflow, review, and escalation owners — 5 min
  2. Set the weekly practice cadence — 5 min
  3. Choose capability, adoption, outcome, and risk signals — 7 min
  4. Define office hours and incident handling — 5 min
  5. Schedule the 30-day decision review — 8 min
Evidence to leave behind

A 30-day playbook with owners, dates, measures, and a decision rule.

Stop condition

Do not call a pilot successful because people used the tool. Compare the agreed evidence with the baseline and recorded risks.

Close and follow-through

The workshop ends when the operating rhythm begins.

  1. Photograph or collect the completed artifacts only with the organization’s approval.
  2. Confirm the final workflow, owner, baseline, success measure, risk measure, and decision date.
  3. Send the clean 30-day playbook within two business days.
  4. Record unresolved issues, incidents, accessibility friction, and changes needed before repetition.
  5. Run the agreed review conversation and record expand, revise, stop, or choose another workflow.
  6. Report observed evidence with its sample and limitations. Do not convert directional change into invented ROI.