Frequently asked questions

Smart Role generates an AI persona from the scenario you describe: a role, a personality, a mood, a hidden agenda and the objections it is allowed to raise. Your team talks to that persona by voice or text, and the conversation is scored against a scorecard you control.

Yes. Voice conversations run in real time with natural turn-taking, and the persona can interrupt or be interrupted. Text mode exists for open spaces and noisy environments, and both modes are evaluated with the same scorecard.

You define the criteria, their weight and what a passing score means. Smart Role scores each criterion and quotes the exact moment in the transcript that justifies the score, so the feedback can be discussed rather than accepted on faith.

Smart Role runs simulations in English, French, Spanish, Italian, Portuguese, German, Dutch, Danish, Swedish, Norwegian, Polish, Romanian, Bulgarian, Greek, Turkish and Arabic. A scenario can be written once and localised per market while keeping the same evaluation criteria.

Under ten minutes for a first scenario. Start from a template, replace the persona and the context with your own, adjust the scorecard, publish. No instructional designer required.

There is a Free plan, with no credit card: 10 role-play sessions with unlimited personas, scenarios and scorecards, every channel and the Zendesk & Intercom integrations. When you're ready, ask for a 30-day free trial of the full platform.

Core is billed per active user: from $29 per active user per month, with unlimited scenarios per user. An active user is anyone who runs at least one role-play during the month, so agents who did not train are not invoiced. Ultimate is priced on custom terms.

Yes. Your bill follows the number of active users each month, and upgrading or downgrading is handled with your account manager, without losing anything you created.

Yes. Registered charities, public education institutions and companies under 10 people can request a discount from the contact page.

All data is stored and processed in the European Union. A Data Processing Agreement is available on every plan, and sub-processors are listed on the GDPR page.

No. Your conversations, scenarios and scorecards are never used to train third-party foundation models, and are never shared with other customers.

Yes. Retention is configurable per workspace, and any user can request deletion of their own recordings and transcripts. Deletion propagates to backups within 30 days.

SAML 2.0 SSO with Okta, Entra ID, Google Workspace and any compliant IdP, plus SCIM provisioning on the Enterprise plan.

Most teams publish their first scenario on day one and roll out to a pilot group the same week. A full department rollout typically takes two to three weeks, mostly spent agreeing on the scorecard.

Yes. Everything in the product is available through a typed API, and webhooks let you push results into your LMS, CRM or data warehouse as they happen.

Support answers in under four business hours in English and French on all plans. Enterprise customers get a named contact and a guaranteed response time.