Live Simulations — role-play with an AI counterpart
Pitch, de-escalate, negotiate and present against an AI that stays in character — then get scored.
- Rehearse high-stakes conversations — sales, escalations, presentations — with zero real-world risk
- Speak your responses and get your delivery (pace, filler words) scored too
- Specific, honest feedback from an AI communication coach after every run
Conversations are a skill you can drill
Live Simulations put you in a real-time role-play with an AI counterpart — an angry customer, a skeptical buyer, a stakeholder you need to move. The brief states your goal; the counterpart stays in character until you end the session.
Type it or say it
You can type every response, but the higher-value rep is speaking: where your browser supports it, tap the mic and dictate. Spoken turns show your words-per-minute and filler count live, and the aggregate delivery across all voice turns is factored into your score — ideal pace is around 110–160 wpm with minimal fillers.
Scoring
When you've had a real exchange, hit End & score. An AI communication coach grades your performance out of 100 against the scenario's rubric — ownership, empathy, clarity, outcomes — and gives specific, actionable feedback. If you answered by voice, expect explicit comments on pace and fillers.
Every run is saved, and your best score per scenario shows on the hub — so re-running a scenario you fumbled is exactly the point.
Tips
- Open strong: state who you are and what the conversation is for.
- Push toward the goal in the brief — the grader scores against it.
- Don't rush the counterpart; acknowledging their position scores better than steamrolling it.
- Re-run scenarios after reading feedback. The second attempt is where the learning lands.
The journey, step by step
| Step | What you do | What you get |
|---|---|---|
| Choose a scenario | Pick a simulation — sales call, support escalation, negotiation, presentation… | You see your goal for the conversation before you start |
| Open the exchange | Type — or tap the mic and speak — your opening line | The AI counterpart responds and stays in character until you end |
| Work the conversation | Push toward your goal turn by turn | Spoken turns show live delivery stats (words per minute, filler count) |
| End & score | Hit End & score after a real exchange | A /100 score with 2–4 sentences of specific, actionable feedback |
| Beat your best | Re-run the scenario | Your best score per simulation is tracked on the hub |
Frequently asked
What kinds of scenarios are there?
Sales calls, support de-escalations, customer-success conversations, negotiations, presentations, written communication, ownership moments and more — each with a category and difficulty.
Does speaking instead of typing change my score?
Yes. Voice turns are analysed for pace and filler words, and the grader judges delivery as well as content — ideal pace is roughly 110–160 words per minute with minimal fillers.
Why can't I end and score immediately?
The grader needs a real exchange to judge. Have at least a couple of genuine turns with the counterpart before you hit End & score.
Who grades the conversation?
An AI communication coach scores your performance (not the counterpart's) out of 100 against the scenario's rubric, rewarding ownership, empathy, clarity and outcomes.
Related guides
Investigate a simulated production incident, resolve it, and get graded like an SRE.
ReadRecord or upload spoken answers; they're transcribed, AI-graded, and reviewable by your mentors.
ReadHow your prompts are scored and improved.
ReadTrack assigned and self-set work with priority, impact and done-ness.
Read