Automated Rep Assessments

• Enablement spent too much time assessing learners, away from resolving performance gaps and producing strategic resources.
• For Highspot, there was potential to decrease satisfaction of our main advocates, eventually leading to churn.

To ship automatic assessments so enablement is free to focus on work with higher impact on win rates.

I led end-to-end UX for the feature:

• collaborating on research and summarizing findings
• setting AI feedback format and quality guidelines
• prototyping flows for learner and enablement personas
• mapping system validations, error states and fallbacks
• collaborating with PM and AI engineering to refine AI output

Working in increments, we were able to ship high quality AI assessments for all submission types, in-lesson feedback for learners and admin controls to fully automate training assessments.

• Reduced submission review time from ~65h to 0h through AI-powered automation.
• Increased learner scores by 9.4 pp on average (88.5% vs. 79.1%) when learners actively revise answers using the AI feedback loop.

Best practices on prompt writing and how to use instructions to decrease the chance of error or hallucinations.