Inspiration
What do you get someone as a gift when the holiday season rolls around? Also, can you really make use of $10 gifts from five different people? We've solved both of these problems with Santatron.
What it does
Santatron allows you to set up a page for that special someone in your life. Other users can then contribute money to a pool for the recipient, as well as suggest keywords relating to the recipient's interests.
The app takes all the suggested keywords, uses a machine learning model to condense them, and uses the most relevant keywords to search Amazon for the perfect gift, within the specified price range.
How we built it
The main web app is built in python using Django. Submitted keywords are recorded in the database, processed through with several Natural Language Processing tools to extract more similar words, and those similar words are ranked based on the similarity scores. The most relevant extracted words are turned into a search query which is sent to the Amazon Data Scraper API. The results from the Amazon search are compared against the price target and displayed to the user.
Challenges we ran into
We had to sort through many iterations of a machine learning model. Even when we discovered an API was the most efficient and error-proof way to implement keyword matching, we still had to sift through many APIs until we found one that met our use case.
Accomplishments that we're proud of
Created the MVP in under 12 hours.
What we learned
How to integrate many complex components together into one object.
What's next for Santatron
- Incorporating contributions using cryptocurrency
- Gift payment through a smart contract, so the gift is automatically ordered
- More optimized search keywords, so the suggestions are better
- User accounts and public/private gift pools
Log in or sign up for Devpost to join the conversation.