We saw a gap in the ever-growing fashion industry where AI could be used to provide programmatic recommendations for a traditionally human concept. Given the difficulty of keeping up with fashion trends, we wanted to not only advise users on fashion choices, but also give them an easy way to actually purchase clothes based on those recommendations.

What it does

Makes fashion recommendations and provides links where you can purchase fashionable items based on any photo of you that you upload to the service.

How we built it

We used the MERN stack and heavily used the Clarifai API to provide recommendations. We also scraped data from multiple fashion sites using a JavaScript crawler to get "inventory" to recommend to users.

Challenges we ran into

Learning new APIs and trying to incorporate as many APIs as possible while still maintaining the spirit of the application.

Accomplishments that we're proud of

Approaching an industry that we were not extremely familiar with at the onset of the project, and produce an idea that we feel could take off in the near future as AI and image recognition tools become more powerful and customizable.

What we learned

An application does not need to have many features for it to provide value to the end-user. If anything, a simple and streamlined user-experience is key to making the application attractive.

What's next for Grace

  • Drop-shipping implementation to garner profits from directing users to aggregated clothing merchandisers
  • Building a more well-defined model for fashion trends and providing better recommendations for obscure clothing
  • Login/user sessions that train a model specific to a user
Share this project: