“Ethics training should be a prerequisite for anyone handling data”

Assia Garbinato, Head of Digital & Innovation at Romande Energie  © Pedro Ribeiro Photography

Assia Garbinato, Head of Digital & Innovation at Romande Energie © Pedro Ribeiro Photography

Assia Garbinato is a data scientist and artificial intelligence expert who completed her PhD at EPFL in 2000. She was appointed Head of Digital & Innovation at Romande Energie in 2021. In this interview, she shares her insights into the potential impact of data on society and explains why the human aspect is central to her philosophy.

What do you remember from your youth in Algeria?

I grew up in Miramar, a coastal district of Algiers. It was a very pleasant place to live, with stunning French colonial-era buildings. We could see straight out over the sea. I have fond memories of that time. I can recall the exceptional light and vivid blue sky. My parents knew how to enjoy life. We had guests over often and our house was a hive of activity.

In my teenage years, things took a turn for the worse with the rise of fundamentalism in Algeria. In 1991, the light gave way to darkness. It all happened in the blink of an eye. I lived through that entire period known as the Dark Decade – with assassinations, a curfew and increasingly oppressive restrictions.

How did that affect your studies and plans for the future?

Getting an education was essential in Algeria – especially for women, since it was our only path to emancipation. So this turn of events only strengthened my resolve. Of course, the years I spent at the Higher National School of Computer Science were a somewhat uneventful period. I couldn’t go out in the evenings and there wasn’t much of a social life.

When you’re trapped in that kind of situation and you can’t see a way out, there’s obviously a temptation to leave in search of a better life. That’s exactly what I did after graduating. I decided to move to Switzerland, and had friends in Lausanne whom I could stay with while I got my bearings.

In 1997, you joined EPFL as a PhD student in computer science. What is it about this field that fascinates you?

My interest in science dates back to when I was a child. My brain is hardwired to seek out logic and meaning. Math was one of my favorite subjects, and I’ve always seen computer science as a fun way to indulge that passion. I continued my studies when I arrived in Switzerland because I knew I wouldn’t be able to pursue my career without a Swiss qualification.

When you graduated from EPFL, was it your intention to work your way up to a senior management position?

Not at all. I never planned out my career. The idea of climbing the corporate ladder didn’t particularly appeal to me. But I knew I wanted to work in industry because I was looking for long-term stability. I felt that would be harder to achieve in a research environment, where PhD students and postdocs tend to come and go.

In 2001, after graduating from EPFL, I joined Lysis, a startup that developed content management software for TV broadcasters. The company was taken over by Kudelski not long after I joined. My first child was born in 2004, at which point I put my career on hold.

Your first management role was at Vaudoises Assurance, where you worked for almost 15 years (2006 to 2019). What did you learn from the experience?

I was entrusted with positions of responsibility quite soon after joining Vaudoises Assurance. I became a team leader in 2008, and then was appointed Head of Data in 2014 and Head of Data & Information Management in 2019. I wasn’t deliberately seeking a promotion to a management role. At the time, I had young children and my only concern was to maintain some sort of work-life balance. But I guess people appreciated the skills and qualities I brought to the table, such as my attention to detail, my ability to see the big picture and my propensity for abstract thought. These are all skills I gained during my time at EPFL.

When you work in insurance, data is your raw material. I drew on my engineering background to build an application with a solid base architecture that could be rolled out organization-wide. My contribution was recognized. I particularly enjoyed working on early machine-learning projects, developing systems for analyzing customer attrition and detecting fraudulent claims.

You were recently appointed as Head of Digital & Innovation at Romande Energie, less than two years after joining the company as Chief Data Officer. How has your role changed now that you’re a senior executive?

It’s a huge responsibility! These days, I have to look at the organization holistically. I’m making and advising on decisions in a much broader range of areas than I’m used to. That requires me to take a step back and look at the bigger picture. I’m constantly planning ahead and making connections. As an action-oriented self-starter who thrives on purpose, I feel incredibly lucky to be doing what I’m doing.

Technology is a wonderful thing, but it isn’t a blank check. We have a duty to use data and algorithms responsibly.

Assia Garbinato

High-profile scandals and leaks can sometimes undermine public trust in how data are managed, and ethical concerns are mounting as technology advances. Do you see way to resolve this conflict?

Technology and ethics are two sides of the same coin. Ethics training should be a prerequisite for anyone handling data. In fact, organizations should go further and draw up binding codes of ethics for data analysts. Technology is a wonderful thing, but it isn’t a blank check. We have a duty to use data and algorithms responsibly.

Social biases in data are a case in point. The only way to address these biases is to have datasets examined by cross-disciplinary teams of experts. There was one scandal where a bank’s mortgage-approval algorithm systematically screened out female applicants. The AI system analyzed the data, identified that there tends to be a gender income gap, and therefore rejected applications from women. In this case, the algorithm itself wasn’t to blame. The problem lay with inherent social biases in raw, uncontextualized data. On a broader level, the digital revolution under way calls for responsible practices. Before using new technology, we have to first assess its social impact. That’s true for its environmental impact, too.

Speaking of the environment, how can AI and data analytics help us use energy more efficiently?

Energy is an inherently intangible and historically cheap commodity. In the past, people rarely gave a second thought to where it came from. But soaring prices, the turmoil in the natural-gas market, the closure of nuclear power plants and the more frequent droughts are shaking up the energy landscape.

Data can play an important role in raising awareness of environmental issues and educating people on their own energy use. Combined with technology, data analytics enables smart resource management. Device charging is an obvious example. Let’s say you plug in your electric car when you get home from work. A data-driven system could delay drawing power from the grid until a later time when peak load has passed.

Data can play an important role in raising awareness of environmental issues and educating people on their own energy use.

Assia Garbinato

In 2018, you took part in EPFL’s very first International Women’s Day event. Women are still a minority in science – especially computer science. Are you noticing any positive changes?

I think it’d be unfair to say that nothing has changed. These days, people are much more aware of equality issues. But women’s underrepresentation in science stems from deep-rooted stereotypes and biases. We need to address these issues early on, in childhood, to understand where the biases come from. It’s not something we can fix overnight. Many women are sorely tempted to give up their careers at certain points in life, such as when they have children or when their work arrangements no longer suit them. They need real support. This will benefit women’s partners, too, because being the main breadwinner can cause a lot of stress and leave people feeling isolated. Striking a better balance in dual-career couples will help to alleviate some of this burden. I sometimes hear people say they’re tired of feminism and equality. But it’s a subject that speaks to our core values. We can’t let it become a fashionable topic that comes and goes with the ebb and flow of societal trends. If we genuinely believe in equality, we have to engage with the subject sincerely and authentically.

Too many young women and girls still overlook science and engineering as viable career paths. What can be done to change this attitude?

There are a number of areas where we can take action. First, it’s impossible to overstate the importance of role models. Children who grow up in settings where women don’t work will find it hard to imagine themselves having a career. The way we raise girls also has a vital role to play here. There’s no escaping the fact that the world of work can be an unforgiving environment. Many women lose confidence or consider abandoning their careers altogether when the going gets tough. That isn’t true to the same extent for men, who are brought up to believe that not working isn’t an option.

There’s no shortage of women engineers in northwest Africa. In fact, around half of students at engineering schools are women. Why? Because they want to build a better life for themselves, and the only way to achieve that is by getting a good job and making a decent living. So they actively seek out well-paid careers – in other words, careers where they’re paid the same as men.

It’s impossible to overstate the importance of role models. Children who grow up in settings where women don’t work will find it hard to imagine themselves having a career.

Assia Garbinato

You’ve spoken at length about the importance of the human aspect in society and in the workplace. Where does this fit into your role as a manager?

Complex 21st-century problems take collective intelligence to solve. Molding a group of individuals into a fully functioning, close-knit team is a discipline in its own right, but it doesn’t get much coverage in conventional business-school programs. Fortunately, concepts such as caring and conscious leadership are increasingly filling this gap. Managing change and leading teams on major projects are part of what being a manager is all about. But you have to do a lot of self-awareness and self-development work before you can reach that stage.

The best managers leave their egos at the door, celebrate team wins and give young talent a chance to shine. And the best teams are those in which every person is able to make an impact.

You seem like an incredibly calm and composed person. Is there anything that irritates you?

Of course! I get unnerved standing still, and I’m a big believer in humility. I’m the kind of person who needs to keep learning and progressing. That’s what makes me feel alive.

Bio

1971 Born in Algiers
1996 Completed a Master’s degree at the Higher National School of Computer Science (ESI) in Algiers
2000 Completed a PhD at EPFL
2019 Appointed Head of Data & Information Management at Vaudoises Assurances
2021 Appointed Head of Digital & Innovation at Romande Energie

In brief

Your favorite spot on the EPFL campus?
The steps under the Esplanade. I love the view across Place Cosandey and beyond.

The best movie you’ve watched recently?
I really enjoyed the Dune remake by Denis Villeneuve.

Your favorite city?
Lausanne. I always associate it with the sense of freedom I felt when I came to Switzerland.

The best time of day?
Early morning. I’m not always awake at that time of day, but I adore the peace and quiet.

Interview published in Dimensions magazine (#4). Electronic version available here.