There are several individuals and funds working on developing Machine Learning based trading software. There are also a myriad Python libraries available for free (open source) that can be used if you have a background in ML and Python programming. It’s a fascinating field.
The new crop of ML trading software is based on Reinforcement Machine Learning (a type of Artificial Intelligence algorithm) in which the developer describes the environment, provides the rules of the system and indicates what the goals are to be achieved. Thereafter, as raw unlabeled data is fed into the system, the AI engine tries various combinations and actions and learns from each of them as it goes along. There is no supervision provided to the software. It continuously improves by itself. Many have claimed that their trading software outperforms humans and while some of these claims may well be true, but their software is not commercially available to retail traders so there is no way to verify their claims first hand.
Reinforcement ML is used in computer games, self driving automobiles and various other applications. A famous example of RML was the AlphaGo software that beat the world’s best Go player. AlphaGo was developed by a UK company called Deep Mind. I believe that Deep Mind invented the concept of Reinforcement Learning or Deep-Q Learning as it is sometimes called.