Ceyla Simeu is Lead Data Scientist at Rexel. In this interview, she tells us about how she grew the confidence to take on public speaking engagements, the challenges she has overcome throughout her career as a black woman in  Data Science, and how increased diversity in the field is essential to make Artificial Intelligence (AI) a force for good.

Interviewed by Anna Marin

Since 2018 you have been working at Rexel, where you started as an intern and are now the Lead Data Scientist. Can you describe what initially sparked your interest in Data Science and tell us about your career journey to your current position?

My interest in data science first started with mathematics. I was born into a family of mathematicians – my parents are both PhD in Computer Science and Automation, so it was very present in my everyday life. When I was a child, my father would often ask me mathematical questions when I was just 5 years old. My parents have really inspired me on my career path, and in many ways, I follow their path.

I went to the University of Grenoble Alpes, where I studied Applied Mathematics and later a Master’s in Data Science. I joined Rexel, my first company, as an intern five years ago. When I started, I had a lot of fresh ideas, as well as good organisation and communication skills, and I think that is why I succeeded in climbing up within the company. I am really grateful that the management team trusted me and believed in me, as I am now the Lead Data Scientist. These communication skills are still integral to my approach to work – as data scientists, we often think we must stay behind our screens but that is not my way of thinking. For me, a data scientist also needs to be a good communicator and be visible.

How have your experiences as a young black woman in STEM shaped your career and personal journey, what challenges have you encountered along the way and how have you overcome them?

It was a long journey to become a lead data scientist. During my Masters, when I was looking for an internship, I came to the difficult realisation that if I removed my picture from my CV, I received more positive responses from different companies. This was tough for me, but I eventually found my internship and got to where I wanted to be, at Rexel.  

I am really integrated at my company, and I think that now, being who I am is a strength. I have a mantra and it is: “Everyone knows me, but no one knows my name”. Because it’s true that when you are black at a big company and in a leadership position - people remember you easily. But I also want people to know my name, and that is what I am working towards.

There is still a long, long way to go for us to be recognised as we are and for the industry to look at skills, rather than looking at skin colour. Of course, there are also stereotypes connected to being a woman, but I think that the tide is thankfully starting to turn in this regard. The battle lies now in bringing about this positive change for black people, particularly black women.

AI is described as the heart of Rexel’s data-driven strategy. Could you give us an insight how you work with AI in your role as a Lead Data Scientist?

Rexel has a data-driven strategy, which means that we have a strategy oriented on data. We have a lot of different types of data on the products we sell, such as transactional data, environmental data and so on, which makes the AI team an integral part of Rexel.

I work with a multi-profile team; we are around 20 people with different types of knowledge. We have data scientists that will build the algorithm and construct the code, but we also have people like software and data engineers because it is important that we can industrialise the code, and business-oriented colleagues who show how our projects bring value to the company and the final user. Thanks to this team our solutions at Rexel are fully automatised, which means that we can use them over and over, saving ourselves work and time in the future. I coordinate the technical implementation of the algorithm and manage the evolution possibilities, and AI is at the centre of this, at the base of the model we use.

As a professional in AI, you're likely aware of the varying public opinions on its use and future applications. Why do you think feelings of fear towards it exist and how do you think it can be mobilised in a positive way?  

AI has been around for a long time, but many people think it is a new trend. We have always used this kind of tool, the difference is that AI is easier for everyone to consume now than it was before. As people, we are often scared about the unknown and I think scepticism is normal. I have been confronted with this within my work, some might be worried that our tool will replace their jobs, and therefore adoption is very important. Adoption enables us to show the user that the tool will be useful to them. That is also why we need people on the team with different profiles, who can explain the tool to diverse business users to reassure them that AI can actually help them become more efficient in their job rather than replace it.

We have also seen an incredible expansion in the use of generative AI in everyday life. I think it will lead to increased productivity, and we need to see this as a tool that will improve the way we work and as an opportunity to develop our jobs. However, we should keep in mind the environmental impact of Generative AI and seek to minimise this – it needs a lot of data, and the number of computations is enormous, which is bad for the environment. Another thing we need to keep in mind is that AI is often biased. Because AI is often based on historical data, it fits the stereotypes of that data. An example of this is the word “nurse” in English. In French the word “nurse” is translated to the word “infirmière” by AI, making the word automatically feminine in French. When AI translates the word “Prime Minister” to French it translates to “premier ministre”, a masculine translation, therefore reinforcing gender stereotypes when it comes to job roles. All of this shows that we will always need humans to validate AI and ensure it is developing in the right direction.  

Continuing with the topic of AI, you recently spoke on the ‘AI for Logistics’ panel at the AI for Industry panel 2023. Can you share insights into the importance of diversity of voices in discussions on AI and technology?

It is true that AI is biased, and sometimes, these kinds of events can be biased too. We often see more men around those round table discussions. So it was great for me to participate in the ‘AI for Logistics’ panel to add my perspective.

I think we need to act early to improve representation; we need to go out to schools and show kids what we do in AI and that we need women in this industry. If we can demystify tech to young girls then this will pave the way for them to become interested in it as a career, and eventually contribute to more representative datasets.

I think we need to act early to improve representation; we need to go out to schools and show kids what we do in AI and that we need women in this industry.

This taps into what you talked about before – the importance of being visible as a data scientist. So, could you share a little bit about what you have learned from these public speaking engagements? How has your experience on the Women Talent Pool Programme contributed to enhancing your confidence in taking on these opportunities?

It has always been my dream to talk about tech in public settings, and the WTP programme has helped me achieve this. In my Career Development sessions, my mentor, Viktorija Smatko-Abaza really encouraged to take on public speaking engagements and because of her and the programme, I accepted to be on the ‘AI for Logistics’ panel and it only made me hungry for more. It made me feel comfortable in my role, in myself and that I deserve to do this. The workshops also helped me to learn about the importance of body language, the importance of networking and effective communication.

It has always been my dream to talk about tech in public settings, and the WTP programme has helped me achieve this.

Finally, do you have any advice for others in data science, just starting their career?

Don’t be afraid to gain visibility within your company. As a data scientist, you don’t always have to be behind your screen. But it is important to know how to structure your code and comment your code. Be aware of the next person who will read your code and help them understand. You also need to learn how to build an algorithm.

Finally, I recommend being full of ideas and try to boost your communication skills. I think this is what will really make the difference on your journey.

Don’t be afraid to gain visibility within your company. As a data scientist, you don’t always have to be behind your screen

Video edited by Claudia Heard