Flyhomes AI Broker
Company
Flyhomes
Tools
Figma
Duration
3 Months
My Role
Product Designer
Feature preview
“What should I look out for with this home?”
Flyhomes AI Broker does home research for you
Homebuyers are able to review a list of call-outs items in the home research report generated by Flyhomes AI Broker.
“Is this home close to high voltage power lines?”
Ask Flyhomes AI Broker anything about homes
Homebuyers are able to ask any questions about homes and get answers immediately.
“Show me comparable analysis.”
Flyhomes AI Broker help you find similar nearby homes sold recently.
Homebuyers are able to view similar nearby homes sold recently to have a better understanding of the home pricing.
Overview
Flyhomes AI Broker is an online AI chat feature designed to enhance the home buying experience.
Business challenge
New agent commission rules intensify market competition
Buyers are now responsible for paying the buyer’s agent commission, and the fee is negotiable, rather than a fixed percentage.
Users’ pain points
How might we provide better service at a lower cost,
while standing out in an increasingly competitive market?
Strengths of Flyhomes
A new product: Flyhomes AI Broker
Align business and user goals with AI solutions
How can we help buyers with AI?
To address this, we need a holistic understanding of the key steps with potential touch points. Among all steps, we identified the ones that have more potentials to be fully or partially completed by AI Broker.
Prioritize features
We interviewed clients and agents to identify valuable AI features for experienced buyers, while collaborating with engineering, data, and legal teams to assess feasibility and risks. This helped us strategically prioritize features across four phases.
Design iteration
After design reviews and usability testing with participants, we went through several design iterations.
Iteration 1: More compact layout and less scrolling
To improve efficiency, I replaced the space-heavy card view with a compact list view, reducing scrolling and making the research report easier to browse—similar to an email inbox layout.
Iteration 2: Easier switch between multiple interested properties
User testing showed buyers often juggle multiple properties and needed an easier way to switch chats. I added a dropdown in the address bar to let them quickly switch between home conversations.
Iteration 3: Individual thread for each call-out item for easier tracking
To improve chat tracking, I replaced the single-thread view with dedicated threads for each call-out item. This allows users to easily follow and revisit specific discussions, rather than losing context in the main chat.
Outcome - Increased 7-day retention rate
We released a basic version of AI Broker with limited features on our current app to gather data and observe behavior. The results were promising—users who engaged with AI Chat showed higher retention rates than those who didn’t.
