Provlepsi

• 

  Website Page

Inspiration

We wanted to tackle food wastage in various businesses while also allowing them to optimize sales.

What it does

Provlepsi allows business to forecast of food sales for various businesses to allow them to prepare for customer demand.

How we built it

We built Provlepsi by combining the frontend Streamlit Python framework designed to show off data-driven machine learning models. Then, we hooked it up to our backend model, which generates the graphs for sales predictions shown in the app. The Pitch deck was made using an online tool called Visme, which is designed for pitch decks.

Challenges we ran into

Some challenges we faced was to find a way to quickly develop a solution that could be hosted and be connected to a model to predict growth. We came across the Streamlit framework and implemented our model with it. Finding data to train our model also proved to be a difficult task as most datasets were simply too small or lacked the parameters we wanted to use.

Accomplishments that we're proud of

We are really proud to present a live web application that is able to demonstrate its potential in predicting user demand for food. Being unfamiliar with these frameworks as well as trying to implement the model within a short time was no easy task. Working efficiently in a team to design, plan, and produce a powerful solution that is now live is amazing!

What we learned

We learned how to use the Streamlit framework, which none of us had such experience with earlier. Also, while we have made applications in the past, the pitch portion of this competition was new to us.

What's next for Provlepsi

In the future, our team will expand Provlepsi to a mobile application, giving it the same functionality that the website currently has. We will also update the app based on user feedback, as was said in the presentation

Built With

• matplotlib
• numpy
• pandas
• python
• streamlit