FortuneCart

Inspiration

Shopping is exciting, but let's be honest...it's SO much better when there's a good sale! Many of us have experienced the frustration of buying a product, only to see it go on sale just days or weeks later. We've all been there. The pain of missing out on those savings hits hard, especially when you don’t have time to return your items. Or worse, when returns aren’t even an option! That's where FortuneCart comes in.

What it does

FortuneCart is your ultimate smart wish list, bringing together all your favourite items from multiple stores into one convenient place. But that's not all, our web app uses predictive machine learning models to forecast the likelihood of the nearest sales, so you can be sure you're getting the best deals on your wishlist items. Plus, we want to encourage mindful shopping practices through our budget tracking feature, where you can set your spending limit and watch your remaining balance update as you mark items as "bought". This way, you can easily track your spending and ensure you’re sticking to your financial goals, while still enjoying the thrill of getting the best deals.

How we built it

We implemented the frontend using HTML, and bootstrap, along with some custom CSS. The backend was built in Python, and we used Flask to communicate between the two. When a user inputs a URL of a product for a new wishlist item, we developed our own algorithm for scraping web data from retail stores to identify the image, price, and name of an item. After scraping all the necessary information, we use the Naive Bayes classifier machine learning model to estimate the nearest sale (compared to today’s date) and the likelihood in which the sale will occur based on sale data from previous years.

Challenges we ran into

The main challenge we faced was retrieving product details from its URL. Not all web pages are created the same so our web scraping strategy isn't effective long term. We also tried looking into different APIs but we have not found relevant ones that can be used without charge during the duration of the hackathon. Due to time constraints, we proceeded with developing our own web scraping algorithm which can retrieve product details from websites with similar page structures, including Zara, H&M and Burberry.

Accomplishments that we're proud of

We're all very proud to have a working final product! Throughout this hackathon, we all experimented with many new technologies and learned a lot so we're very happy that all the different components we worked on came together, and that we were able to bring our idea to life.

What we learned

We learned a lot about product development, from turning an idea into a pitch to building a working product. Our team also gained hands-on experience with Python-based web development, which was new to some of us. This project helped us grow both technically and creatively as we worked through the challenges of building FortuneCart.

What's next for FortuneCart

We're hoping to add a tag and filtering feature, where users can add tags to their wishlist items and filter through them. We also want to add an option for users to scan their closet, and keep inventory of all the items they already own so that we can notify users when items in their Wishlist are similar to items they already own. Lastly, we want to provide the opportunity for companies to partner with FortuneCart, as an additional tool that can be used to market upcoming sales.