We aim to be a resource for those who want to be active in their investments but don't necessarily have the bandwidth to constantly keep up with all the news. So we try to find some signal in the noise and learn what kind of news articles tend to lead to large swings in price. We do this through a combination of machine learning and signal processing and we bias towards having higher precision on the PR curve to make each message count. When we come across an article that we think may be relevant, we kick it over to the user through a SMS message to make any final decisions. We use data from the Bloomberg API for both training and test time.

Share this project: