As people participating to many hackathons over the past 4 years, we have always told ourselves "we should definitely continue this project after the hackathon". However, many other things started becoming priorities, and we got lost in university projects, and internships, and we were never able, or incentivised, to follow through with our projects until the end. This is a problem through which many university students, and other people, are going though at some point in their lives, but our lives are such fast paced and distracted, that we do not get the chance to follow through with projects and ideas that could, potentially, have a significant benefit on the world.

What it does

StartNet is supposed to breach many of the gaps that aspiring entrepreneurs are facing, in order to generate a market where more independent organisations are able to compete and develop, providing value to everyone involved in the process. There are 3 principal stakeholders we are considering: the Inventor, who is presenting an idea, with a pitch and relevant tags. This process automatically makes you an inventor account. The project, alongside projects proposed by other peers, will be accumulated in the UI, putting the most similar/relevant projects nearby, so that different entrepreneurs could collaborate if their fields coincide, or if they could be completed by the other's expertise. The builders are people interested in supporting a startup. They can join, and they will be shown projects which are close in scope with their own area of interest. The investors are the final category of stakeholders, which can use the app if they want to invest in a new company, with novel ideas. The investors can be searched for by the Inventors, or they can search themselves among the ideas closest to their interest.

How we built it

Our project is a web app, built with the front-end in REACT and the backend in FLASK and Heroku. The similarity between the projects is generated using NLP, by matching tags based on how similar they are to each other (fields of synonyms and word2vec).

Challenges we ran into

Making a bubble chart is a boss battle NLP is even harder Making a fully fledged app with many user journeys in 24 hours (at least there was Redbull)

Accomplishments that we're proud of

The entire app is almost fully functional! We are amazed of our newly found front-end development skills, as well as NLP related programming.

What we learned

NLP React Heroku Flask SQLite MUI for React

Share this project: