HomePilot AI

AI for Good: A demo that defines a real-world problem outside of college life and shows the AI addressing it.

Comment

HomePilot AI

🎬 Demo Video:
https://homepilotai-demo.vercel.app/

About the Project

Inspiration

Access to affordable housing remains one of the most challenging real-world problems individuals and families face today. Many people do not know whether they can realistically afford to rent or buy, what housing assistance programs they qualify for, or where to begin their search.

We created HomePilot AI to help simplify this process.

HomePilot AI acts as an intelligent affordability assistant that evaluates a user's financial situation, identifies support programs they may qualify for, and recommends realistic housing options. The goal is to provide structured guidance that helps users make informed housing decisions with confidence.

This directly supports the AI for Good challenge by addressing housing accessibility through practical, data-driven assistance.

What it Does

HomePilot AI is an AI-powered housing affordability assistant that helps users:

  • Estimate affordable rent ranges
  • Estimate affordable home purchase budgets
  • Analyze financial readiness signals
  • Match with housing assistance programs
  • Discover properties aligned with their budget
  • Receive personalized recommendation rankings

Instead of searching blindly across multiple platforms, users receive structured recommendations tailored to their financial profile.

Example affordability logic used:

$$ Affordable\ Rent \approx 0.30 \times Monthly\ Income $$

This estimate is combined with location preferences and eligibility insights to produce smarter recommendations.

How We Built It

We designed HomePilot AI as a modular full-stack application using a scalable architecture.

Frontend

The frontend was built using Flutter, providing:

  • Authentication screens
  • Profile setup workflow
  • Dashboard affordability insights
  • Property recommendation views
  • Grant eligibility display
  • Saved listings tracking

Flutter enabled rapid cross-platform development while maintaining a clean UI structure.

Backend

The backend was built using Java Spring Boot, including:

  • REST API architecture
  • JWT authentication
  • PostgreSQL-ready schema structure
  • Service-layer architecture
  • Repository-layer architecture

This design allows the system to scale toward real-world integrations.

AI Agents

HomePilot AI includes modular service agents:

  • Affordability Agent
  • Grant Matching Agent
  • Property Recommendation Agent
  • Mortgage Estimation Agent

Each agent analyzes user inputs and produces structured decision-support insights.

The architecture supports future integrations with:

  • Google Maps API
  • Zillow API
  • Realtor API
  • Plaid API
  • Vertex AI / Gemini

Challenges We Ran Into

One challenge was designing a system that balances simplicity for users with meaningful financial insight.

Housing affordability depends on multiple variables:

  • Income range
  • Employment status
  • Household size
  • Credit estimate
  • Geographic differences
  • Program eligibility rules

To manage this complexity, we separated logic into specialized AI service agents so each component could remain modular and interpretable.

Another challenge was designing a recommendation scoring pipeline that evaluates listings beyond price alone. We addressed this by combining:

  • Affordability compatibility
  • Preference alignment
  • Assistance eligibility matching

What We Learned

During this project we learned how to:

  • Design modular AI service architectures
  • Connect Flutter interfaces with Spring Boot APIs
  • Build recommendation scoring pipelines
  • Translate financial signals into user-friendly insights
  • Structure scalable MVP systems for real-world expansion

Most importantly, we learned how AI can support decision-making systems that improve access to essential resources like housing.

What's Next for HomePilot AI

Future versions of HomePilot AI could include:

  • Live property listing integrations
  • Bank-ready mortgage estimators
  • Location-based affordability heatmaps
  • Plaid-powered financial readiness insights
  • Google Maps commute-aware recommendations
  • Vertex AI-powered personalization models

Our long-term vision is to evolve HomePilot AI into a practical housing navigation assistant that helps people move from uncertainty to ownership readiness.

Built With

Share this project:

Updates