Build a practice tool for sales calls with AI buyers
A go-to-market leader at Replit with no coding background used Replit Agent to build a simulator where reps practise cold and discovery calls against AI buyer personas and get scored.
What they did
In his first two weeks on the job he described the tool to the agent in natural language and had it running in under two hours. The simulator scores reps against sales frameworks such as MEDDIC or Sandler and gives coaching notes after each call.
What happened
He expects it to shorten new rep ramp-up from about three months to under two weeks.
Keep in mind
The ramp-up figure is an expectation, not a measured result, and the story is published by Replit about its own employee.
Try it with your agent
Build a web app where a salesperson can practise calls with an AI buyer. Let me pick a persona from <your buyer types> and a call type such as cold call or discovery. After the call, score the rep against <your sales framework> and list three specific things to improve, quoting the moments from the transcript.
Replace the <angle brackets> with your own details.