DTU_FOOD_APP

Inspiration -

When you go to a food outlet in DTU and have to wait for a good 15-20 mins to get it prepared, why not order it directly from your class or anywhere else and have it ready when you reach ?

What it does

Gives a medium for buyers and vendors to do online purchase and transactions for food specifically in DTU oulets such as Deltech, Bistro Cafe, etc. It also has a smart recommender system using vector embeddings semantic search using a lightweight sentence transformer model.

How we built it

Started of with the backend -- first created a menu for dtu with all the outlets , their menus and sample reviews, stringified the entire information for each food item and generated a vector embedding, then made a UI for both vendor and buyer, and integrated the 2.

Challenges we ran into

Couldnt add a review system and automated tag updation for food items based on the reviews given by users.

Accomplishments that we're proud of

Completed most of the features I intended to achieve in the given timeframe.

What we learned

Streamlit skills, how to actually use a model from hugging face

What's next for DTU_FOOD_APP

The feature that I couldnt add earlier -- adding reviews and updating embeddings for food items iteratively.

Built With

• python
• streamlit