Housing Stability 'System navigator'

Empowering case managers and families with an intelligent reasoning engine that decodes program eligibility and maps human referrals.

Comment

Inspiration

Inspiration

Navigating public assistance programs during a housing crisis is a bureaucratic nightmare. Vulnerable individuals facing eviction or homelessness are forced to parse complex state and federal policy guidelines, calculated income thresholds, and hidden compliance requirements while under severe emotional distress. We built the Housing Stability 'System Navigator' to eliminate this cognitive friction. Inspired by the need for accessible social infrastructure, our goal was to build a compassionate, transparent portal that instantly translates intimidating eligibility math into clear, actionable next steps.

How We Built It

We designed and engineered the platform using an agile, low-code accelerated stack:

  • Frontend UI/UX: Developed using HTML5, CSS3, and JavaScript, leveraging premium open-source icons via Lucide to keep the interface highly accessible and stress-free.
  • Rapid Prototyping: Powered by Lovable.dev alongside Claude 3.5 Sonnet and Gemini acting as pair-programmers to rapidly iterate the responsive intake form and terminal logs.
  • AI Reasoning Engine: Built a rule-based simulation environment that ingests a user's multi-variable context matrix (such as household size $H$ and gross income $I$) to cross-reference localized assistance policies without handling real-world PII.

Challenges We Faced

Our biggest technical hurdle was building a real-time, simulated terminal interface that streams logic logs back to the user transparently. We had to ensure that parsing asynchronous console sequences didn't freeze the browser UI. Additionally, modeling the overlapping categorical criteria (such as veteran flags, disability indicators, and immediate shelter crises) required meticulously structuring our decision routing matrix to prevent contradictory program recommendations.

What We Learned

This project reinforced the power of transparent AI. We discovered that when dealing with high-stakes social services, simply displaying a "qualified/not qualified" badge is not enough—users need to see the why behind the logic to feel empowered. We also deeply explored the principles of Responsible AI, learning how to engineer strict human-in-the-loop guardrails and dynamic referral systems to combat user over-reliance on automated logic tools.

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for Housing Stability 'System navigator'

Built With

Share this project:

Updates