Project Overview

AI Role Play

AI Role Play — real-time practice call interface
COMPANY
Highspot
TEAM
1 designer, 1 PM, 4 engs
TIMELINE
2023-25

Role plays have been a useful tool for sales teams. But in 2025, adding AI into the mix became a common request.

Problem

  • For customers, there was a missing training type becoming a new industry standard.
  • For Highspot, role play was a new requirement for sales and existing accounts were having higher risk of churn.

Goal

To deliver an AI-driven practice environment that simulates real sales calls and drive rep skill growth.

My Role

I led the design strategy and execution, including:

  • leveraging design system primitives for real-time call environments
  • prototyping scenario setup workflows and prompt auto-fill flows
  • establishing new org-wide UX patterns for editing AI outputs
  • refining post-call AI assessment reports

What Shipped

After shipping the initial MVP, we refined interactions for the call experience and expanded creation, assessment and role play related AI conversations.

Results

  • Role play has become the #1 practice method in our platform.
  • We contributed to a +131% growth in domains using AI licenses (from 45 to 115).

Favorite Learning

Emerging standards for creation flows as part of AI conversations, and best practices for Human In The Loop interactions.

Linking creation inputs to scenario prompts.
Linking creation inputs to scenario prompts.
User journey planning.
User journey planning.
A summary of visual explorations.
A summary of visual explorations.
Workflow documentation.
Workflow documentation.
Role play ecosystem: item viewer, call intro, call controls and assessment.
Role play ecosystem: item viewer, call intro, call controls and assessment.
Agentic creation of Role Play.
Agentic creation of Role Play.
Manual creation flow.
Manual creation flow.

Need more details?

Email me for a full case study walkthrough: