Inspiration

The idea of making personalization seem natural was what motivated us. We wanted consumers to simply express their preferences through pictures rather than through long surveys or forms, allowing images to speak for themselves.

What it does

An interactive web application called Customify displays collections of photographs with a theme. Our machine learning algorithm analyzes these visual selections to deduce personality traits, preferences, and product affinities after users select the ones that most appeal to them. After that, it suggests customized products from Frasers Group companies including Flannels and Sports Direct.

How we built it

For image-based interaction, we created a simple, user-friendly front end and linked it to a backend that was fueled by real-time data gathering and processing. In order to find associations between product categories and visual preferences, our machine learning model analyzes selection patterns.

Challenges we ran into

Assuring consistent performance with dynamic data, training precise models from sparse starting samples, and creating a user interface (UI) that seems engaging and natural while gathering valuable data were our biggest hurdles.

Accomplishments that we're proud of

We are proud in having developed a functional system that seamlessly combines technology, psychology, and design to turn a straightforward selection procedure into a customized suggestion engine.

What we learned

We discovered how data-driven modeling and user experience design can produce surprisingly accurate customisation, and how visual clues may be potent predictors of preference.

What's next for Customify

In order to make product discovery more intelligent, quick, and individualized than before, we then intend to grow our dataset, improve the machine learning algorithms, and incorporate Customify straight into Frasers Group's e-commerce ecosystem.

Share this project:

Updates