Realtor develops a mobile app to discover the value of a property in the street

Machine Learning Real estate

In search of the perfect house

Finding the perfect property is never easy. Even if a potential client is actively searching for a property to buy, he does not necessarily have all the information at hand when he needs it.

The right house might be just in front of you, and yet...

This article was written inspired by the original story published here.
Let's assume a potential customer passes in front of a lovely neighbourhood and a certain property catches his attention. Best case scenario, the property is for sale and there is a sign in front of it with the name and the phone number of the real estate agency that sells it. But even if the potential customer wanted to know more about the property, he would still need to stop doing what he is doing, call the agency (or visit its website) and find information on it. Many sales opportunities could be lost just because the customer was too busy doing other stuff and when he finally got home, he forgot to search for that specific property


What if there was a magical mobile app?

Could you imagine an app capable of knowing the price and availability of a property in the street by simply pointing your camera at it?


Machine Learning Real estate


Fortunately, we live in the future and there is no need for magic, just a little bit of technology can help us here..


Creative and mobile solutions for the busy customers.


Realtor, one of the largest real estate companies in the USA, decided to develop a mobile app that uses your GPS coordinates to know exactly your position and narrow down a list of the properties you could potentially be interested in.

Using Cloud Vision, an amazing image recognition API from Google Cloud, the address of the properties can be extracted via Optical Character Recognition (OCR). With this information, the app can fetch and display the property data from Realtor's database and display it in augmented reality for the user to see.


Ok, so how does it work?


Google is able to achieve this Machine Learning trick, as OCR is one of the many things that are automatically delivered when Cloud Vision processes an image with text on it.

True... the application could be improved so that it does not rely on the address, but that it could actually compare images from the street to the images in the real estate database. For that, however, we would need to do supervised learning to train a classification model, probably in TensorFlow.

But as they say, "Perfect is the enemy of good" :-) so I suppose this could be a part of the V2.0 of the application.


MLab, the Machine Learning specialists at your service!


If Machine Learning inspires you and you think you would like to implement a use case in your organisation, please contact us. We are independant and we will recommend and integrate the technology that adapts the best to your needs. If available technologies don't satisfy your needs, we can always develop a custom model tailored to your project.

Best of all, you would be surprised to learn the implementation cost of such a solution :-)


DisclaimerMLab was not directly involved in the development of this project. We simply publish this case study as a source of inspiration on what Machine Learning can achieve.