Your people talk with an AI character who has a role, a goal and real reasons to push back. They speak or type the way they would on the job. Every attempt comes back scored on each skill, with each score linked to the moment in the conversation that earned it.
What we really hang our hat on here is our extraordinary customer service.
Honestly, that's what every insurance person says. Can you give me something more concrete than just “extraordinary service”?
Why she pushed back: the character pushes back when the learner makes a general claim about better coverage or service without concrete, policy-level detail.
A practice conversation is a realistic exchange with an AI character who has a role, a goal and reasons to resist. The learner speaks or types as they would on the job. myQuest scores every skill in the scenario from 1 to 5 and links each score to the moment in the conversation that earned it, so you can see who is ready.
An internal auditor course added practice conversations. Learners took the same final exam as the traditional course.
The AI simulation felt unfamiliar at first, but after a few tries, I really got into it. The practice was extremely valuable because, ultimately, an auditor has to perform in the field, and practice is what helps you feel prepared.
Every practice conversation, from a claims call to a performance review, is built from the same five parts and scored with the same method.
The situation, the learner's role and a goal you can observe, like “confirm a specific next step before the call ends.”
A persona, an opening line, a voice and a personality type, plus private context the learner never sees.
Up to five moments of resistance, each with what sets it off, the real worry behind it, and what resolves it.
The skills this conversation is scored on, from your framework or the myQuest library.
Language and translations, voice or text-only, and the character's accent.
A hurdle is a moment of resistance with something that sets it off, a real worry behind it, and a way through. Miss the worry and the character pushes back. Handle it and the conversation moves. You set how proactive the character is and how hard the hurdles are to clear.
Whether the goal was met, what worked, what to do next time, and a 1 to 5 score on each skill. Skills with several parts get a score for each part. This is a real attempt at the Harborline renewal call (a fictional insurance agency).
You recognized that Dana was feeling pressure from the 18% increase and tried to explore whether anything besides price mattered to her.
Turn uncovered risk into a specific, low-pressure next step, such as a 20-minute call to review the policy together.
“Dana repeatedly stated an explicit objection: she wanted clear, concrete details about what she was paying for.”
How skill mapping works →Start from 16 ready-made personality types, from hostile to cooperative. The Doubter needs proof and logic before moving forward. The Speedster wants the bottom line. The Worrier needs reassurance and simple explanations. Or describe your own.
Learners speak by default. Any practice conversation can run text-only, and any learner can choose text for themselves. Write a practice conversation once, translate it into the languages your teams work in, and set an accent for the character's voice.
Describe the conversation and AI drafts the scenario, character and hurdles. Your experts then check every field, so the practice conversation sounds like your business, not a script. See how building works →
Yes, the practice partner is an AI character. The difference is what comes out: a 1 to 5 score on each skill, linked to the moment in the conversation that earned it, so practice tells you who is ready.
Most take 2 to 7 minutes, short enough to practice between calls or shifts.
Yes. Any practice conversation can be set to text-only, and any learner can choose text for themselves.
Each character has a persona, a personality type, private context and up to five hurdles that are set off by what the learner says or misses.
Yes. Describe the conversation and AI drafts it, or write it from scratch. Every field can be edited before it goes live.
Many. Write a practice conversation in one language and translate it for each team, or run it in the learner's own language.
A typical pilot runs 30 days. You'll work with a myQuest instructional designer who runs your pilot, and we'll decide together what success looks like.