top of page
Housing Connector AI Assistant
Designing an AI-assisted mobile experience to help renters navigate housing search, applications, communication, and support resources.

Develop for Good

Housing Connector

May-Aug 2026

Figma

Team-Based

PROJECT SNAPSHOT

A real product problem, not just a screen redesign.

Role
Product/UX Designer

Client
Housing Connector

Program

Develop for Good

Timeline
May-Aug 2026

OVERVIEW

Making an already difficult housing process easier to navigate.

Housing Connector is a tech-for-good nonprofit that helps people access stable housing. Our team explored how an AI assistant could reduce usability barriers within Renter Hub and provide renters with more consistent support throughout the housing search and application process.

The Problem

  • Renters struggled to find housing that met their needs.

  • Application requirements could be confusing.

  • Communication gaps could delay progress.

  • Support was needed across multiple stages of the renter journey.

My Contribution

  • UX research synthesis

  • User flows and wireframes

  • Interactive Figma prototyping

  • Usability testing and iteration

  • UI components and design-system work

  • Cross-functional collaboration

RESEARCH

Understanding where renters get stuck.

6
Housing platforms analyzed

9

 


Existing research resources reviewed

Interviews

Case manager perspectives on renter challenges
Survey

Additional feedback on housing-search barriers
KEY INSIGHTS

Four findings shaped the product direction.

01 · Finding suitable housing is difficult

71% of renters reported difficulty finding housing that met their needs on the platform.

02 · Applications can be hard to understand

Application language, requirements, and varying levels of digital literacy created friction.

03 · Renters need consistent guidance

Communication gaps could leave renters without centralized support at important moments.

04 · AI can supplement human support

Case managers showed openness to AI when it was approachable, transparent, and supportive.

UX STRUCTURE

Turning research into a product flow.

We mapped the information architecture and user journey before detailed UI work so the assistant could support renters at meaningful points throughout the process.

02_information_architecture.png

Information Architecture

User Flow

AI EXPERIENCE

Designing AI people could understand and trust.

Personality + Tone

  • Clear and direct

  • Approachable and supportive

  • Transparent about limitations

  • Human-centered without pretending to be human

 

Guardrails

  • Escalate to human staff when needed

  • Simplify housing terminology

  • Never predict approval or rejection

  • Avoid making decisions on a renter's behalf

  • Do not hallucinate information

04_ai_branding_guidelines.png
05_ai_rules_guardrails.png
DESIGN EVOLUTION

From low-fidelity concepts to an interactive prototype.

Lo-fi → Mid-fi → Prototype

​

We explored layouts and AI features collaboratively, evaluated concepts against user needs, and then refined them into connected flows with typography, UI components, brand styling, and interactive behavior.

1789560973438-c638669f-a1cf-4523-847b-a7af4df375d4_8.jpg
06_mid_fidelity_designs.png
USABILITY TESTING

Testing exposed what still wasn't working.

14
User Tests

3
Major feedback themes

~75 %
Found the experience/features helpful

V1
Handoff target

Hidden interactions

Some controls, pinned-message behavior, and checklist interactions were not easy to discover.

Confusing language

Some terminology and prompts needed clearer, more specific wording.

Missing preferences

Users wanted more detailed options for amenities, housing type, and accessibility needs.

07_user_testing_screens.jpg
FINAL PRODUCT

The final experience.

01

Guided Onboarding

Introduces renters to the assistant, explains what it can help with, and sets expectations before they begin.

09_final_designs_overview.jpg
10_onboarding_saved_homes.jpg
11_onboarding_application_guide.jpg
12_onboarding_assistant_intro.jpg
15_housing_search_property_detail_edited
13_housing_search_questionnaire.jpg
14_housing_search_recommendations.jpg
02

Personalized Housing Search
Uses renter preferences and eligibility information to surface relevant housing options and AI-supported summaries.

03

Guided Applications

Supports application requirements, document uploads, progress tracking, and dedicated conversations.

15_housing_search_property_detail_edited
13_housing_search_questionnaire.jpg
14_housing_search_recommendations.jpg
18_guided_application_ai_summary.jpg
17_guided_application_document_upload.jpg
16_guided_application_checklist.jpg
04

Communication Support
Notification nudges and AI-assisted email drafting help reduce communication gaps between renters, property managers, and case managers.

05

Resources + FAQS
Centralizes guides, forms, FAQS, and additional support resources when users need help beyond the assistant.

21_communication_email_draft.png
20_communication_email_template.png
19_communication_notification.png
DESIGN SYSTEM + IMPACT

Designed  to be consistent, reusable, and ready for implementation.

70+
Screens

180+
Components

14
User Tests

12.5K+

Current Renter Hub users potentially reached
image.png
image.png
"These designs are gold!"

Jan Van Bruggen · Lead Engineer, Housing Connector
Reflection

What I learned.

AI needs boundaries

Designing the assistant meant defining both what it should do and what it should never do.
Research should stay visible

The strongest product decisions could be traced back to renter and case-manager pain points.
Handoff changes design thinking

Working toward implementation pushed the team to think beyond individual screens and design reusable patterns.
See the interactive experience.
bottom of page