Inspiration
- Chef Gordon Ramsay and Infinite Craft
What it does
- Analyzes recipes and ensures that all selected food preferences are matched for each recipe.
How we built it
- We built Ramsay using Python. We use Flask to connect the frontend and backend. We used Gemini AI to analyze the recipes. We used the Featherless AI chat model Meta-Llama-3.1-8B-Instruct for a chatbot that the user can interact with and ask questions about the recipes.
Challenges we ran into
- We tried using Google's Programmable Search Engine, but faced issues and had to switch to standard Python web scraping based on the site's search results. We also had trouble getting the Gemini AI to work. Also, one of our team members got locked out of his car and was stuck at another event.
Accomplishments that we're proud of
- We got it working quickly and then just started adding features to it. We worked strongly as a team and stayed focused.
What we learned
- We learned how to use an LLM in our code to analyze and validate information. We learned how to incorporate an AI chatbot for the user to use and made sure it has the relevant page information for the chat.
What's next for Ramsay
- Quicker responses and a trained agent specifically built for analyzing recipes.
Built With
- allrecipes
- beautiful-soup
- cloudscraper
- css
- featherless
- flask
- gemini
- html
- javascript
- python
- vertex
Log in or sign up for Devpost to join the conversation.