Inspiration
SillyArg is a play on the word "Sillage," the trail of scent left in the air by perfume. The two of us often struggle to figure out what scents we like beyond the common ones found in stores like Sephora. So, we decided to create an app that helps people shortlist the top 5 related scents to those that can be commonly found. Our goal is to make discovering new fragrances easier and more enjoyable, helping users explore options they might never have considered while expanding their scent horizons.
What it does
Our App has 3 key features:
Perfume Database: We web scraped all the perfumes we could from Sephora, and displayed it all on our webpage, together with its fragrance qualities.
Specifying Notes: When it comes to perfumes, deciding what notes matter the most to you affects your decision. This is why we allow users to pick if they want to find perfumes most similar to the top, middle or bottom notes.
AI Integration: We integrated Gemini to retrieve the Top 5 most popular scents that matches the specified notes and fragrance qualities of the perfume selected.
How we built it
This app was built with Vite + React with a FastAPI backend. We used BeautifulSoup and Selenium to Web Scrape, MongoDB to store data, and Gemini as our AI model.
Challenges we ran into
Spent a lot of time trying to scrape data from Sephora accurately, as well as making an intuitive but simple user interface.
Accomplishments that we're proud of
It is Xin Yi's first hackathon! First time web scraping and using Vite.
What we learned
Everything we built we learnt during the 24 hours
What's next for SillyArg
More data and possibly accounts to store the generated content. We can probably also refine the prompt for the model a lot more!
Built With
- beautiful-soup
- fastapi
- gemini
- javascript
- python
- react
- selenium
- tailwindcss
- vite
Log in or sign up for Devpost to join the conversation.