AI and Big Data to Reveal Real-time Car Choices

The Background

Scouring the internet for different products can be a pain. There’s just so much information and so many products, that it takes lots of time and effort to find exactly what you’re looking for. Out of personal frustration from unfruitful searches, the founder and CEO of Find.me Stefan Ceulemans decided to create a digital platform that could aggregate all available data about a specific product and then provide recommendations to users based on their profiles. Find.me streamlines the registration process and provides a complete overview of all of the products that the market offers. On one hand, it gives customers a faster and more complete offer and on the other, it results in more efficient advertising spending for retailers. It’s a win win situation. Partnering with us, Find.me developed its first successful digital product find.me/my car, a platform that helps people who already had some preferences and even those who have no idea what type of car they would like to have.

The Challenge

Find.me needed a vendor that could help them develop a digital platform for their find.me/my car service. Driven by a lack of IT expert capacity and high cost in its domestic market in Belgium, they started to sort after nearshore development companies. The ideal vendor had to be able to seamlessly collaborate with Find.me’s team (a Data Processing Algorithms developer and Data Scientist located in Belgium and the CTO in Panama). After shortlisting several candidates, meeting and evaluating all of them, we seemed to be the best fit for Find.me’s big project.

Strategic Partnership

Find.me has been working with us since 2017. At first, Find.me tested pwrteams by asking for a trial phase whereby our team had to put together several mockups and then consult with Find.me on the technological stack for further development. As soon as the scope had been agreed with, our team designed the interface and started building out the solution. We helped to build Find.me’s Data Science model right from scratch. The model contained a mixture of analytic approaches for Find.me/my car:

• Analysis and processing of static, content-based models;

• Dynamic, collaborative filtering models for individual items, based on user actions and preferences;

• A rule-based model that combined collaborative subcategories and content-based models.

Our team comprised of a development team lead, a full-stack developer for the backend, and a data engineer. While our company operates remotely from Ukraine, the client points out that “it feels like they’re literally in the same room as us.”

The Results

At present, the basic platform for car search with virtual assistance is complete and successfully launched on the market. Find.me/my car helps people to easily find the best option for their future car and it saves them tons of hours they would have needed to spend on the internet. They get a personalised offer, based on a proprietary developed matching algorithm.

Who benefits from using find.me/car:

People who already have a car preference. They can find their ideal car using the given filtering option which allows them to choose different categories like the car brand, model and price among others.

People who have no idea what kind of car they would like. They have the unique option to fill in some information about themselves, their lifestyle, driving preferences, the stylistics of the car they want, and even more. Find.me’s innovative virtual assistance will help match them to the ideal car.