Inspiration

This project was inspired by vending machines and a Cohere workshop I attended. I wanted to play around with LLMs and I decided to make a product prototype for a futuristic time where we can materialize food out of thin air.

What it does

When the python script is run with the Ardiuno connected to a port, the script will run a sequence asking you for your meal choices. Each choice has two options to choose from. These options are chosen from a trained Cohere AI module (4 individual modules total). The first choice is the type of cuisine for the meal. The second is for the main course. The third is for the dessert. Afterword, the price of the items are calculated (with tax) and a receipt is 'dispensed' (virtually).

How we built it

I built it in Python using the knowledge of Cohere I gained from their workshop.

Challenges we ran into

Figuring out how to get a consistent output from the LLM was difficult (not to mention actually learning how to use the API), I ended up training individual modules to improve their outputs. It's still not perfect, but it's a lot better than before. Trying to get Python and my Ardiuno to communicate both ways using the serial monitor was very difficult - if I manage to get it to work I'll edit this

Accomplishments that we're proud of

I'm very happy that I was able to actually somewhat complete and submit a project!

What we learned

I learned a lot about LLMs and how to use them

What's next for Restaurant Worldwide!

Bug fixing, physical input/output improvement, and larger training datasets

Built With

Share this project:

Updates