Dog Park Homes Finder

Inspiration

I recently relocated from Oklahoma to Washington with my dog. During the home search process, I realized that while there are many tools to evaluate homes based on schools, commute times, or other amenities, there is almost nothing designed specifically for dog owners.

One of my biggest pain points was figuring out which homes were close to good dog parks. I often had to manually search Google Maps, read park reviews, and estimate distances from potential homes. That experience inspired me to build DogParkHomesFinder, a tool that helps people discover homes near highly rated dog parks.

What it does

DogParkHomesFinder is a full-stack web application that helps users find homes for sale near great dog parks. Users can enter a natural-language query such as “Seattle, walkable parks under $900k”, and the application returns homes located near highly rated dog parks.

The results are displayed on an interactive map along with nearby parks, park ratings, and insights derived from park reviews. This allows users to quickly understand how dog-friendly a neighborhood is when evaluating potential homes.

How we built it

The project is built as a full-stack application with a React + TypeScript frontend and a Spring Boot backend.

The frontend uses Google Maps to visualize homes and dog parks on an interactive map. The backend orchestrates several services:

• Amazon Nova 2 Lite (via AWS Bedrock) parses natural-language user queries into structured search filters such as location and radius.
• Google Places API is used to find dog parks in the requested location and retrieve park reviews.
• Amazon Nova analyzes those reviews to produce structured scores for aspects such as cleanliness, parking, and dog-friendliness.
• RentCast API provides home listings near selected dog parks.
• Redis caching stores park analysis results to avoid repeated AI calls and improve performance.

Together, these components create a pipeline that transforms a simple user query into nearby park discovery, park quality analysis, and relevant home listings.

Challenges we ran into

One challenge was integrating multiple services—AI models, mapping APIs, and housing data—into a single coherent workflow. Each component produced different types of data that needed to be normalized and combined.

Another challenge was designing a meaningful way to evaluate dog parks. Reviews are unstructured text, so using AI to extract useful signals from those reviews required experimentation.

Finally, as with most hackathons, we had to carefully balance feature scope with the limited time available.

Accomplishments that we're proud of

We are proud that we built a working end-to-end system that combines AI, geospatial search, and real-world housing data. Users can type a natural-language query and immediately see homes near high-quality dog parks on a map.

We also successfully integrated Amazon Nova to both interpret user intent and analyze dog park reviews, turning unstructured text into structured insights that help users compare parks.

Most importantly, the project solves a real problem we personally experienced while relocating with our dog.

What we learned

This project taught us how to combine AI services, external APIs, and geospatial data into a practical application. We also learned how AI models like Amazon Nova can be used not just for chat interfaces, but for structured tasks such as query parsing and review analysis.

Working under hackathon constraints also reinforced the importance of focusing on the core user problem and building a simple but functional prototype quickly.

What's next for Dog Park Homes Finder

There are several directions we would like to explore next:

• Expanding the scoring system to include additional signals such as park size, amenities, and trail access.
• Improving the natural-language search capabilities to support more complex filters.
• Incorporating additional dog-friendly amenities such as vets, pet stores, and dog-friendly cafés.
• Expanding coverage to support more cities and real estate markets.

Ultimately, we want DogParkHomesFinder to help dog owners find not just a home, but a neighborhood that both they and their dogs will love.

Built With

• amazon-bedrock
• amazon-nova
• eslint
• google-maps-javascript-api
• google-places
• java
• maven
• react
• redis
• rentcast-api
• springboot
• vite