I started to work as a software developer in 2016 at VisoTech. I had just received my Ph.D. and was looking for opportunities to work in business and for solving real-world problems. By luck, I ended up developing software for energy companies and did my first steps in the field of trading. I as well got a glimpse of financial engineering, risk management, and automated power and gas trading. Roughly the year after I started to look into machine learning and, in particular, deep learning in application to price forecasting.
I joined enspired because the idea behind the company was to take automated energy trading to a completely new level. It appealed to me that instead of writing trading software for the customers, we would do the trading for them. This of course gives us full decision power for implementing our own ideas and for doing the necessary research in the area. Four months after joining I must say it completely paid off.
I believe there are still a lot of opportunities in the energy market for automated trading. Compared to the financial markets, the intraday energy market still falls behind concerning the number of order executions, order execution speed, number of trading participants, the number of automated trading solutions, as well as the number of automated trading algorithms utilizing machine learning. In the financial world, there is serious competition between the research groups, quants, and traders to develop models predicting stock prices with high accuracy. Portfolio optimization is another research area based on advances in machine learning. I remember reading about a program called “Star”, which was operating on a financial exchange and learned its stock-picking strategy on its own by scanning a dozen of different price tendencies (see Dark Pools by Scott Patterson).
I am certain that we are witnessing a transition from manual to completely automated electronic trading in the energy market right now similar to what happened in the financial sector 10-20 years ago. And the use of machine learning plays a major role in this transition. Finding trading signals in the enormous amount of trading data, weather data, load data is a task for machines to crunch. We, as humans, can draw only very few conclusions from the data we see, and we easily miss important trends. We are as well prone to emotional risks, panics, fatigue, and just simply typo errors. Machines are free of those and machines can learn on their own.
I search for work inspiration online as well as offline while chatting with my colleagues. I am blessed to be accompanied by bright people, who ask difficult questions. Online is the next important resource. I like to read blogs about programming, as the topics of these blogs give one a clue about the progress or the trends in the field of computer science. From my scientific background, it was natural that I would be interested in the field of data science. And so, a few years later there is a lot of information around on the internet, as the field is growing very rapidly.