Engineering my own career path
How Engineer Michael switched from software to quant.
At IMC, success is about playing to your strengths. We try to match people with the work (and team) that fits their individual skills and interests. So, when graduate Software Engineer Michael demonstrated an interest in machine learning, the stage was set for his switch to one of IMC’s research teams as a Quant Software Engineer. In this article, Michael explains how the switch has played out, and how IMC life is treating him so far.
What brought you to IMC?
I studied electrical and computer systems engineering at university. IMC appealed to me for two main reasons. The first was the company’s reputation for harnessing and developing cutting-edge technology solutions. This was evident throughout the application process: at times, it felt like I was learning more from my interviewers than I did in some of my university lectures!
Reason number two was the great work culture. This also shone through during the interviews – and it’s been the case throughout my time at IMC so far.
How did you end up moving to an engineering role on a machine learning team?
After completing IMC’s Global Traineeship, I joined the Japan Execution desk in the Sydney office, which focuses on latency, exchange connectivity and ensuring that IMC’s systems are fast and reliable enough to make the trades we want.
However, I have had an interest in machine learning since completing my honours project on deep learning. Upon joining IMC, I heard from the research team about the unique constraints of using machine learning in algorithmic trading, and the impressive engineering solutions required to use low-latency machine learning in production.
When I was offered the opportunity a few months after starting to join the Base Valuation Algorithm Development team – a research-focused part of the business responsible for implementing new valuation strategies for equities and futures – I was eager to take it!
The team uses cutting-edge technology and it’s also a key engine for IMC’s trading of Delta One products (a class of financial instruments that includes shares and exchange-traded funds). Delta One trading is an exciting, fast-growing area, which made the opportunity even more attractive.
What was it like transitioning to a new team at this early stage?
It’s been a steep learning curve. Not everything has changed: as a Quant Software Engineer, I’m still very focused on coding. However, where I previously focused more on the individual bytes we send to the exchange, I now spend more time implementing trading-related algorithms and developing IMC’s systems to support our research pipeline.
The environment I’m working in is also different. For one thing, the ‘Base Val Algorithm’ team has a global focus and works closely with various teams across IMC, both trading and technology. I’ve found myself working much more closely with other teams at IMC to try and achieve these diverse goals.
This requires broad knowledge of IMC’s systems and trading styles. My background in electrical engineering has definitely come in handy. IMC is a technology-driven company at its core, so having a strong understanding of how our network systems and hardware work is a great advantage.
What does a typical day look like now?
I get to the office around 9.30 am, usually after a quick gym session. My role involves working with various team members across operational trading, research and data engineering. So, at the start of the day, I liaise with the particular trading desk or engineering team I’m working with.
The rest of the day is usually spent looking at ways to improve our trading systems, which currently involves two main focuses. The first is improving the workflow of our research teams by improving the quantity and quality of data we have available for analysis and accelerating the training of our machine learning models. This empowers our research team to constantly generate new ideas for better trading. A strong research pipeline enables my second focus: taking ideas from research for new ways of trading and implementing them in high-performance C++. Low latency is critical at IMC, even in machine learning, and my team ensures that our systems perform well in even the busiest market conditions.
How have you adjusted to life at IMC overall?
Life at IMC is fun and engaging. The social calendar is always packed, whether it’s team dinners, a new outdoor activity or the weekend away. While moving to a new city can be difficult, Sydney’s amazing beaches (and weather) are the perfect cure for homesickness. There’s no excuse not to get out there and meet new people.
And work certainly keeps me busy. There’s always some new challenge or complexity to get my head around – and plenty of smart, interesting people to keep me on my toes!
Engineers – whether it’s software or hardware – are natural problem-solvers. So it’s the perfect environment for people like me: things are constantly changing and it’s up to us to find a creative way to get to the answer. It’s certainly never boring!
Do you love problem-solving? Think you have what it takes to be a software or hardware engineer at IMC?