Inspiration

After looking for a new place to stay for school we noticed how dull and overwhelmingly complicated it was to easily search through properties. Searching through real estate should be fun and engaging!

  • Real estate software can be clunky and overwhelming to use
  • You need to know what you are looking for going into it, as there isn't good exploration and user feedback tools for discovering new properties.

What it does

Gives back top matching homes/condos in your area and provides a simple way to search through them. The idea is to scale it up to a full real estate service that allows users to connect with real estate agents.

Core concepts

  • Smart results to suggest houses you might like based on swiping history (using AI)
  • Fast queries for many listings using BigQuery and BigQueryML
  • Easy to use UI that makes it easy and fun to find your next property and discover what you like (by saving data to your profile)

How we built it

We decided on Moleculer microservices backend with a React-Native frontend as they both allow for ease of development and future scalability. Python was setup to analyze large data models so we can get feedback on what a user likes. BigQuery was used so that we can scale up to large datasets and provide fast results within a large array of parameters (including geolocation). BigQuery also helps with ML analysis on large housing datasets.

We used Docker with a Mongodb image for local development. Deployments using gCloud and heroku.

Challenges we ran into

The time constraint made it tough to get a functioning app off the ground especially only with two people on the team. Also coordinating who works on what and at what time was definitely a learning curve.

Accomplishments that we're proud of

The swiping feature was a big moment for us, neither of us thought we were going to pull it off going into the competition. Getting the app working e2e with real data served based on your location from BigQuery.

What we learned

How to make large and insightful queries using BigQuery. How to deploy demo versions of expo snacks.

What's next for Meant To Be

Connect the full machine learning pipeline to give back recommended results to the user.

Possibly pitching the app to real estate companies and seeing if it is something they would like to acquire.

Share this project:

Updates