The aim of this 10-days project was to answer a business need from BPI France data team, that needed an MVP to build upon that would be used to classify start-ups between 'deeptech' and 'non-deeptech'. Such classification was defined by BPI France team.
The project relies principally on data from Dealroom API as well as Google Patents data.
We took on developing and training a logistic regression model to perform such classification. We also built a simple web-app (MVP) with Streamlit, and the code was then deployed on Heroku.
The resulting MVP is accessible here: https://bpideeptechdemo.herokuapp.com/
This project is only presented here for pedagogical purpose and to display a 10-days long data science project. It was the result of a voluntary collaboration between students at Le Wagon Data Science Bootcamp and BPI France.