Twitter sentiment analysis app backend, hosted on Heroku.
The app provides a dashboard which can give a weeks analysis of the the public sentiment of your search-term, according to Twitter.
It is able to run on either HuggingFace's Transformers package, and NLTK's Vader algorithm. The deployed version uses Vader, as Transformers' filesize are too large for the free tier on Heroku.
The classifier is wrapped in a Flask web application, providing an API end-point for the front-end interface.
Below are some gifs of it in action.