I built a simulator where three AI customer stakeholders debate an AI sales rep in real time. Then it grades the AI rep’s performance and coaches them on what to do next. See the 40-sec demo below.
IT wants security. Finance wants ROI. Travel wants ease of use. They interrupt each other, push back, and argue while the AI rep tries to get them all to yes.
This is a tough scenario by design. A group demo where competing priorities clash. In real life, reps sometimes get 1:1 time with each stakeholder. But training for the hardest room makes the easier ones feel manageable.
The human rep doesn’t participate. They watch, observe, and learn. It’s sales training where your team sees how messaging lands against realistic buyer dynamics without the pressure of a live deal.
Run it twice and you get a completely different conversation. Every response is a live AI call, not a script.
Here’s the starting prompt:
“Please build a simulator where an AI sales rep presents to a buying committee of 3 AI stakeholders based on the attached strategy document. They should react, interrupt, ask tough questions, and debate each other.
Every response should be a live AI call, not scripted, so each run produces a different conversation. I want to watch, not participate.
At the end, show me which stakeholders are sold, which aren’t, and why. Then grade the AI rep’s performance and give coaching: what worked, what they missed, and suggested next steps.”
Your strategy doc does the heavy lifting. The personas, objection handling, and value props you already built become the AI’s playbook. Garbage in, garbage out.
The first version won’t be perfect. Review what it builds and refine from there. Sometimes that means adjusting the prompt. Sometimes it means improving your strategy doc. Either way, you learn something useful.
Try the simulator yourself and read the full breakdown in my latest newsletter. Links in the comments. Share with us what you think.