Skip to main content Skip to secondary navigation

Data analytics platform for rental property investors

Main content start

Team members
(Pictured left to right): Jorge Armenta, Galahad Mai, Shreya Mantha, Samantha Moadel

Faculty mentor
Prof. Markus Pelger

Sponsor organization
Lumina is a property tech startup that manages rental properties on behalf of real estate investors.

Project description

Across most destination short-term rentals, travelers are often forced to choose between overly-priced rigid hotels or unreliable hosts, which can lead to the following disappointing experiences:

  • Most short-term rental marketplaces connect hosts and guests. But, there is a huge issue with reliability and property management.
  • Commercialized short-term rentals are frequently operated without supporting surrounding communities and integrating guests into local communities. This can have a negative impact on small communities abroad.
  • Commercialized short-term rental guests are often met with inflated rental costs, yet have no source of comparison for affordable prices.

As Lumina specializes their platform to service American investors, they want to develop a platform that leverages rental property data to offer real-time financial information, investment yield, and a dashboard interface for investors to manage their properties through the platform. Investors will also be able to discover potential rental property investment opportunities.

The goal for this project was to develop a minimum-viable product for this new platform while operationalizing and streamlining the data analysis engine that powers its features. In the end, Lumina was able to test their hypothesis around the business need for such a platform and expand their product portfolio as they scale.

Techniques and solutions

For this project, we developed a predictive model for property price and occupancy rate using an available dataset and provided a comprehensive evaluation of the model’s parameters and their relative importance. We built 4 k-nearest neighbors (KNN) models for this purpose. While we considered other model types, including linear regression, random forest, and gradient boosting, we decided upon KNN.

These models will help property owners and Lumina partners select the best properties in terms of profitability and revenue estimates that will yield increasingly personalized, practical, and affordable destination travel experiences.

Presentation video

Embed Code

2023 senior projects