ML engineers are in high demand – and the reasons are clear. Artificial intelligence, of which ML is one of the core techniques, offers apparently infinite potential to simplify and enhance tasks usually performed by humans, such as text understanding and generation, speech recognition, and image processing. For those who’re technically inclined and have a background in computer science or software programming, this opens up an exciting range of new career opportunities.
But what exactly does the role entail at a wealth manager like Julius Baer? “I would phrase it as transforming ideas into innovations,” says Dominik. “We apply AI technology to the bank’s data to find ways of enhancing our daily work.” Recently, Dominik has been working with some of the bank’s own data, such as publications, to make it usable for large language models. “We’re teaching the model the Julius Baer ‘lingo’.”
Dominik believes the value of his work lies in reducing routine tasks so that colleagues can focus on more value-adding activities. “In wealth management, especially at Julius Baer, personal relationships are central,” says Dominik, “and that’s an area where AI will never be able to outdo humans. But it does offer the potential to enhance these relationships by enriching the interaction and decision basis, for example by retrieving, preparing, or summarising, data.”
A prize-winning introduction to Julius Baer
Dominik’s fascination with AI evolved out of his teenage interest in video editing. He says he was particularly excited by GenAI’s potential when it came to text to image processing, particularly the creations of media artist, Refik Anadol, who has also partnered with Julius Baer as part of the bank’s NEXT initiative. “He’s an idol of mine. I asked my Data Sciences professor at university if it was possible to do something similar as part of my course and she taught me the fundamentals.” Dominik made the move to Julius Baer after winning Julius Baer’s annual Best Master’s Thesis in Business Information Technology prize at ZHAW School of Management and Law. “It was clearly meant to be,” he says. “When I won the prize, my grandmother told me that my uncle’s godfather was former Executive Board member Fritz Laager, who served alongside former Chairman Hans J. Bär.”
Dominik’s route into machine learning highlights the different career paths open to computer programmers and data scientists. “As I was more interested in working with my hands, I began my professional career with an apprenticeship with a rail manufacturing company and obtained a Federal Vocational Baccalaureate (Berufsmatura) before enrolling in university.” He believes this broadened his view of the working world, even if banking was still uncharted territory for him when he joined. “I was surprised to discover that, like any other company, there are so many different departments with different functions.”
Using curiosity for career growth
Perhaps unsurprisingly for somebody with a scientific interest, Dominik’s fellow ML engineer Marine took an empirical route into her role. “My first year at university was largely a case of ‘try it and see’. During her studies in mathematics and computer science, she chanced upon machine learning when a group of tech-savvy friends began studying it and enjoyed the hands-on, minds-on approach. “What attracted me about ML is that it’s very applied. You need maths to understand the theory, but programming to implement it.”
Marine initially arrived at Julius Baer by applying for a working student position and meanwhile moved into a permanent role. She is currently working on a use case for one of the bank’s units using large language models. “Basically, we’re taking the data created by our Investment Research team and train the models behind our in-house GenAI tool with it, in order to help users find answers to their questions more easily.” Naturally curious, Marine is clearly somebody who spends a lot of time seeking out answers to her own questions. “When I hear in the media about a new technology, I want to find out how it works. I dive into the research papers behind it to understand how it’s possible and what’s happening in the background.”
Traditionally, women have been underrepresented in STEM (Science, Technology, Engineering, and Mathematics) careers – Marine says fewer than ten per cent of her fellow Data Science classmates at ETH Zurich were women. She believes that giving girls greater exposure to technical and engineering subjects at an early age – such as the Julius Baer’s annual Kids4IT day in Switzerland – could help to spark their interest. “I think giving young people the chance to try out computer science in high school could attract more women to roles like this.”
Early strategic choices reap rewards
Paul took a more targeted route into the role. “The university where I studied for my bachelor’s degree in biomedical engineering didn’t offer any machine learning courses at that time, so I had to study extracurricularly. I enrolled in online courses in machine learning and then did a master’s degree in AI at the University of Zurich to learn more about the theoretical, underlying concepts.”
Paul’s responsibility in the team is to orchestrate the landscape of machine learning and large language models used by the bank and make sure that they run reliably and quickly. Despite growing up near Germany’s banking capital Frankfurt, like the others the world of banking was new territory for him. What motivated him to join Julius Baer was the bank’s strategic focus on leveraging the potential of AI as an early mover in this area. “Julius Baer employs very smart people in wealth management and I can make their lives easier by developing tools that let them focus on their core business,” he says, adding that, “Thanks to the decisions the bank has made in recent years in this area, the applications we’re working on are advanced and close to the state-of-the-art in their technical implementation.”
He puts this advantage down to the Julius Baer’s decision to host its hardware on its own premises. “By keeping data in-house, we’ve opened up innovative and safe ways for data utilisation. Since the data remains within our secure environment, it’s protected from external access. It was a smart decision to work in this way. Now it’s on my colleagues and I to help ensure we keep that competitive advantage, or ideally increase it!”
A culture of openness and experimentation
Although their role revolves around designing algorithms that allow machines to function without direct human assistance, all three ML engineers cite working together as one of their most important job motivations. “It’s always fun coming in to work beside the team,” says Dominik. Communication skills are a key part of an ML engineer’s skillset because, ultimately, they need to be able to collaborate with experts from different areas in order to make a machine learning model that serves the bank’s purpose-driven, client-focused approach. “The relationship managers are interested and open enough to let us explain how we can help them and to explain their daily tasks to us,” says Dominik. “It’s a win-win.”
True to her inclination to “try things out and see”, Marine says that, in a field like AI and machine learning, the bank’s emphasis on a bottom-up culture is also essential. “AI is a really fast field – things are changing constantly. To foster innovation, it’s important to give people the space to experiment. A top-down approach would be too slow for us in order to stay ahead.”
Julius Baer offers opportunities to grow‚ outside of the box
This openness to experimentation also brings benefits for the three ML engineers in terms of their own opportunities for development, especially in such a fledgling field. “Even within our own team, we have flexibility to shape our own careers,” says Paul. “We can choose to go down the computer science route where we can dive into cutting-edge computational techniques, or the data science career path, exploring how data distributions, probability, and statistical analysis can reveal new insights. The doors are also open for us to rotate into different teams, such as the investment or governance teams, to broaden our horizons.”
According to the World Economic Forum, an impressive 97 million new roles are projected to emerge by 2025 as a result of AI. Looking beyond the now, what does the future hold for those pursuing a career in this frontier-busting field? “For ML engineers like us, this is likely to create opportunities as employers seek a combination of people with the business insights to understand what the market needs and how they can best leverage advancements in machine learning.”
Dominik believes this tradition of embracing new technology is simply part of the inevitable march of progress. “AI is something that at some point in the future every company will embrace.” He’s excited about the technology’s potential. “It’s not just facilitating the lives of humans in the workplace but also on a really advanced and broader scale, through autonomous driving and flying or maintaining infrastructure – all things that are going to have a real impact on human lives.”