Proptech
Conversational AI
BuildingLink AI Concierge
Designing an assistant that helps residents get things done, fast
One conversational front door for residents, with less manual work for staff
BuildingLink helps property managers, developers, and condominium boards across the globe deliver a superior resident experience while streamlining maintenance and operations. As part of a small product team, I helped design an AI concierge chatbot built into BuildingLink's resident app, giving residents a single conversational front door for the tasks they used to dig through menus to find: paying rent, reserving an amenity, requesting maintenance, getting a parking permit, and much more.
Research
User research uncovered where residents got stuck between wanting help and finding the right menu item, so I mapped the highest-friction tasks (payments, amenities, maintenance, parking) as the chatbot's first conversational flows.
Ideation and wireframing
I sketched multiple ways residents might ask for what they needed, wireframing conversational flows and menu layouts side-by-side to find the right balance between guided quick-actions and open conversation.
Systematic prototyping
Leveraging BuildingLink's existing design system, I prototyped the assistant's visual language, including message bubbles, task cards, and quick-reply menus, while iterating closely with product and engineering to keep every screen technically feasible.
QA and GTM
Through structured usability testing and a phased rollout, I refined edge cases in the conversation flow and helped prepare the launch plan for property managers and residents alike.
I leaned on AI-assisted design tools throughout ideation, wireframing, and prototyping, spinning up and comparing layout directions, rough wireframes, polished mockups, and user-flow diagrams in a fraction of the usual time. That meant structural decisions got made against real options instead of a single point of view, and left more room for the usability testing and edge-case work that actually needed a human.
Problem
Residents buried in menus
Finding the right form for a routine task meant digging through several layers of the app — most residents gave up and called the front desk instead.
Solution
One conversational front door
Residents can now just say what they need. The assistant surfaces the right task — payment, amenity, maintenance, parking — in one message instead of a multi-tap menu hunt.
[Image: user-flow diagram, old vs. new]
Problem
A one-size-fits-all task list
The original flat list of options treated every resident the same, regardless of what they actually tend to ask for.
Solution
A menu that flexes to how residents ask
I explored and tested multiple layouts: a scannable icon grid, and options grouped by intent (requests vs. amenities & community) so the assistant's structure matches how residents actually think about their to-dos.
Problem
Repair requests missing the details maintenance actually needs
A resident filling out a generic maintenance form often skips the specifics: what's actually wrong, how urgent it is, etc. This leaves maintenance staff to call back searching for more answers before they can even schedule the job. Even worse, the maintenance staff shows up unprepared.
Solution
A guided conversation that gets it right the first time
The assistant walks residents through category, a plain-language description, and urgency in a few taps. Then confirms their contact info before submitting, so maintenance receives a complete, accurate request without a single callback.
Problem
Booking an amenity meant guessing at availability
Residents browsing a static amenity list had no idea what was actually open: they'd pick a date, submit a request, and wait to hear back before finding out whether it worked
Solution
Real-time availability, in plain language
The assistant shows photos, hours, location, and capacity right in the conversation, understands a request like "this Saturday afternoon," and only offers the time slots that are actually free so residents land on a confirmed booking in a few taps instead of the trial and error of submitting and waiting.
Problem
High-volume, low-complexity front desk calls
Front desk staff spent a large share of their day fielding the same handful of routine requests (parking permits, payment questions, amenity bookings, to name a few) that didn't need a human.
Solution
Self-serve for the routine stuff
The chatbot resolves the repetitive, low-complexity requests instantly, freeing staff to focus on issues that actually need a person.
“Our front desk used to spend half the day walking residents through the same requests over and over. Now residents just ask the assistant and it's done. We've gotten real time back to focus on things that actually need a person.”
Property manager in Miami
Key measurements
35% decrease in front desk call volume
68% of routine requests resolved via self-service on first attempt
41% increase in resident app engagement
Boosting user engagement, cutting support hours, and driving recurring revenue