Skip to content
Pondara Pondara docs

Run an engagement

An engagement is one pass through one scenario. You pick it, you read it, you act through the hybrid dialogue, you reflect, and your trainer renders a verdict. This guide walks you through it end to end.

Before you begin

  • A scenario has been assigned to you (or you’ve picked one from the scenario library yourself).
  • You’ve read the situation. Going in cold is fine — the AI plays the other side, not a coach — but the situation matters.

Step 1 — Pick a scenario

Open your home or the scenario library. Each scenario shows:

  • the field (e.g. nursing, hospitality, sales)
  • the actors you’ll encounter (the customer, the patient, the supervisor)
  • the dimensions it exercises (which of the eight sub-competencies this scenario will draw evidence in)

Read the situation before you press Begin engagement.

Step 2 — Interpret

The first phase asks you to name what you see in the situation — what matters, what’s missing, what’s at stake. Type a few sentences. There is no right answer; your interpretation is your starting point.

Step 3 — Act through the hybrid dialogue

The AI plays the in-scenario actor (customer, patient, supervisor). You act as yourself. The AI responds in role, not as a guide — it will not tell you the right answer.

Type as you would speak. Slang is fine. Hesitation is fine. Whatever you say becomes part of the material your trainer reviews.

See the AI as sparring partner for why the AI behaves this way.

Step 4 — Reflect

When the engagement ends, the AI sparring partner opens reflective prompts. Answer in your own words. The reflection is part of the record; see reflect after action for how to do it well.

Step 5 — Read the evidence

After reflection you’ll see two things on each dimension the scenario exercised:

  • the AI signal — what the sparring partner observed in the dialogue
  • the trainer verdict — accepted, adjusted, or rejected

Both are visible. The trainer’s line is canonical — it’s the only one that moves your competency record. See evidence and verdicts for how the two combine.

When something goes wrong

  • You can’t find the scenario. Ask your trainer — assignments are cohort-scoped and they may not have promoted it yet.
  • The dialogue stalls or the page reloads. Your progress is saved on every turn. Refresh and click Continue from your home.
  • The AI says something unrealistic. Note it for your trainer or use the in-dialogue feedback control. Field experts use that signal to tune the scenario.