Using Machine Learning to pick stocks

Designing a digital platform prototype for better stock picks


The Challenge

Creating a platform that can understand and interpret both positive and negative sentiment, around a stock.

The Outcome

An intelligent platform prototype, that can pick stocks that make a real return.

A 53% profit in

one month.

When picking stocks, traditionally investors will look at the P&L of a company, the Board members and their experience. Investors will also look at what others are saying about a stock online.

Reading all of this information takes time. So, we built a prototype platform called Buffly, that allows investors to pick a stock based on what positive or negative sentiment is being written about that stock, in that very moment.

The platform reads and understands comments on social media posted about a particular stock, and gives the investor a read out of whether good or bad things are being said about it.

This allows the investor to quickly surmise how the market feels about a particular stock, so they can make more informed investment decisions in less time.

In testing, on a £10K investment the platform delivered a profit of £5316, in one month.