Prompt Coach: Designing the Human Side of AI
Context - Before AI UX Was a Job Title
Before "AI UX" was a job title, I was designing it. Prompt Coach was born from a simple insight: AI is only as useful as the interface that guides people into it. Built in 3 days at the 2023 Compass Hackathon, it won the #MOVEFAST Award and shipped to production — one of the earliest examples of human-centered prompt design in a real estate product.
The Problem - The Missing Piece Wasn't the AI
Real estate agents needed personalized, platform-ready marketing copy, but generic AI output wasn't cutting it. The missing piece wasn't the AI. It was the UX layer that helped users give the AI what it needed to produce something genuinely useful.
What We Built - The Interface Did the Prompting Invisibly
Prompt Coach guided agents through a structured workflow, auto-populating MLS listing data, defining tone, audience, and brand voice, so that by the time AI generated their copy, it actually sounded like them. The interface did the prompting work for the user, invisibly.
The Team - 72 Hours, Six People, One Shipped Feature
Led UX and UI across a cross-functional team of six, including an Agent Experience Manager and front-end and back-end engineers, going from concept to production-ready feature in 72 hours.
Outcome - Award-Winning, Shipped, and Used in Production
#MOVEFAST Award winner, 2023 Compass Hackathon. Shipped to production and used by agents.
#MOVEFAST Award winner — 2023 Compass Hackathon
Entry point: agents land here to begin building their marketing copy
Step 1: What is the listing for? Agents define the context before auto-populating MLS data
Step 2: Listing details auto-populated from MLS with one click
Step 3: Agents add a description to personalize the copy further
Step 4: Exisiting brand
Step 5: Agents describe their brand voice and audience
Step 6: Preview generated copy before sending
Final output: personalized, platform-ready marketing copy
