A worked example

My SME gave me 74 slides. Now what?

Follow a training intake and training needs analysis example, from an SME content dump to a recommendation you could take into a stakeholder meeting.

Fictional teaching example. The deck, evidence, agent suggestions and review below are scripted illustrations. This is not a customer case study, a recorded product run, or a claim of measured results.

The request

“Turn this product deck into a course for the new sales reps. We need it by Friday.”

The stakeholder wants better discovery conversations. The proposed solution arrives before anyone checks what is getting in the way.

Starting material: 74 slides, a deadline, and a reported behavior gap.

The useful output

A scope recommendation that distinguishes practice from reference material, names missing evidence, and asks for a specific decision.

What changes: the conversation moves from covering every slide to agreeing what reps need to do and what is worth testing.

What remains unknown: the cause of the gap and whether the proposed response will help.

The decisions behind the output

Work through the request.

Open each step to inspect the reasoning. These are example instructions you can adapt in the Agents workspace; they do not run AI on this page.

1. Scout: clarify the behavior and missing evidence

Request: “The stakeholder says new reps pitch features before understanding the customer. What should we clarify before choosing a course?”

Illustrative contribution: Identify the discovery behaviors, examine permissioned call samples, compare experienced and new reps, and check workflow and incentives. A slide deck cannot establish the cause.

Your decision: Treat the stakeholder's explanation as a hypothesis. Ask for evidence before approving the full build.

2. Nova: sort the material without mistaking content for evidence

The invented deck contains 28 product-reference slides, 18 procedure slides, 12 discovery examples, 10 repeated or outdated slides, and 6 policy slides. Total: 74.

  • Reference: make product facts findable when needed.
  • Workflow support: validate whether procedures belong in a task aid.
  • Practice: use discovery examples to draft decision situations.
  • Review: ask the SME to confirm what can be removed or updated.
  • Policy: obtain the policy owner's approval before changing mandatory material.

Your decision: Use this as an inventory. It does not prove a skill gap or authorize cutting policy content.

3. Practitioner review: reject an impractical suggestion

Initial scripted suggestion: “Add weekly manager coaching to reinforce discovery skills.”

That assumes managers have time for weekly coaching. In this example they do not. Revise the plan without relying on capacity we have not secured.

Revised example: Pilot two self-contained practice attempts with feedback. Seek one realism review from a manager, subject to availability.

Your decision: Keep the practice idea, remove the recurring coaching dependency, and record the constraint in shared project memory. This is what human review adds.

4. Luma and Echo: propose a small response worth testing

If the call review supports the hypothesis, draft a discovery-question aid and two situations: a customer who asks for a price immediately, and a customer who requests a feature without explaining the underlying need.

Each practice attempt asks the rep to choose a discovery question, explain the choice, and review feedback. Confirm realism and accessibility before use.

Your decision: By Friday, propose a validated scope and drafts. Do not promise finished training before capacity and approvals are confirmed.

5. Axiom: prepare the recommendation and decision request

Compare a full deck conversion, a reference guide, and a small practice pilot. Make the recommendation conditional on the evidence and name its limits.

Your decision: Ask the stakeholder to approve evidence review, nominate reviewers, and agree a pilot scope. Save the rationale so the next specialist starts with the same context.

Read the assembled stakeholder brief

The finished work sample

A recommendation you can inspect.

Fictional stakeholder brief. No outcomes have been measured. Download an editable plain-text copy to adapt the structure.

Discovery conversations: scope recommendation

Requested solution
Convert a 74-slide product deck into a course for new sales reps by Friday.
Reported problem
New reps reportedly pitch features before asking discovery questions. The cause is not established.
Evidence still needed
Review permissioned call samples against agreed behaviors. Compare new and experienced reps; examine workflow, incentives, and available information. Validate mandatory material with its policy owner.
Options
A full course preserves coverage but may miss the problem. A reference guide supports lookup but may not develop discovery skills. A question aid plus practice pilot targets the reported behavior, subject to diagnosis.
Provisional recommendation
If the call review supports a skill gap, pilot a discovery-question aid and two practice situations. Reuse product facts as reference, validate procedures for workflow support, and review outdated or mandatory content with its owner.
Constraint and human judgment
Managers have no protected weekly coaching time in this example. Remove that dependency; request a one-time realism review only if capacity is confirmed.
Friday scope
A validated scope, a draft question aid, and two storyboarded practice situations. Confirm reviewer availability before committing.
How to evaluate
Agree observable discovery criteria before the pilot. Compare consistently selected pre/post call samples using the same rubric. Report sample size, reviewer agreement and other changes. Do not claim revenue impact from practice completion.
Decision requested
Approve evidence review and pilot scoping. Nominate a sales reviewer and policy owner, confirm availability, and agree the review date.

Bring your own request

Start with three better questions.

You do not need a 74-slide deck. Bring a short, non-confidential description of a request you are working on.

  1. Sign in to the Agents workspace and create a project. Membership access is required.
  2. Add the requested deliverable, intended behavior, deadline, evidence and constraints to the brief.
  3. Choose Scout and ask: “Separate supplied evidence from assumptions. Give me three questions to clarify before I accept this assignment. Do not invent findings.”
  4. Review the response. Keep, edit or reject the questions, and save your reasoning in project memory.

Your first useful output: three questions you can take into the next conversation. Check whether they reveal a gap you actually need to resolve.

The public example is free to read and download. The live workspace generates a fresh response from your brief; it will differ from this scripted walkthrough. Use only information you are authorized to share. Privacy Policy