On 4 May 2026 Modar Sulaiman defended his doctoral dissertation “From Data to Fair Decisions: On Ensuring Fairness in Machine Learning Models“ (supervisor Kallol Roy). On this occasion, we spoke with Modar about his path in research and how his journey in academia has developed.

At what moment did you realize that a PhD would not be what you initially expected?

I think I started realizing it in the first month of my PhD, but it became much clearer by the end of the first semester. As I began to reflect more deeply on what the work actually required, I noticed a gap between my initial expectations and the reality of the research, which helped me adjust my understanding of the PhD journey.

If you had to describe your PhD in one sentence today, what would it be?

A journey of understanding how to balance fairness and performance in artificial intelligence.

What were you feeling on your very first day as a PhD student?

On my first day as a PhD student at the University of Tartu, I felt a strong sense of excitement and motivation, similar to how I usually approach new academic environments. I saw it as a meaningful challenge and a responsibility I was committed to taking on. However, the start was also unusual due to the COVID-19 pandemic, which created uncertainty about studying in a new country and how the coming years would unfold. At that time, I was unsure how long the situation would last, but things gradually improved after 2020, allowing the PhD journey to become more stable and structured.

What is the biggest misconception people have about doing a PhD?

The biggest misconception about a PhD is that it is mainly about doing research and writing a thesis to eventually earn a degree. It is a much broader developmental process. It involves learning how to think differently, how to approach open-ended problems, and how to develop new ways of tackling research questions. It also requires resilience, as progress is often non-linear, and the ability to stay motivated and rebuild confidence after setbacks and failures.

How did you first discover the research topics that eventually led you to your PhD?

I think it started from gradually building knowledge around the problem of AI bias. Moreover, I read widely across different areas such as mathematics, physics, philosophy, and computer science, which I also genuinely enjoyed. This diverse background helped me connect ideas more easily and develop a broader understanding of what I was reading. Over time, I began to see how some concepts relate to each other and how they could be used to address research questions. I also believe that research contains, to some extent, a process of discovery, where ideas and answers already exist in some form, and one of the researcher’s roles is to connect the dots and uncover them.

Was there a specific moment or problem that made you think: this is what I want to work on?

Yes, there was a specific moment in the latter part of my PhD studies when I worked on the problem of learning with noisy labels. I became interested in how addressing this type of challenge could also be relevant to the broader issue of AI bias. That connection helped me see more clearly the direction I wanted to pursue in my research at the end of my PhD studies.

Why did Estonia feel like the right place to pursue this journey? 

Estonia has its strong reputation in IT innovation and its high-quality educational system. Despite its small population, it is a remarkable example of what a country can achieve in a relatively short time since independence in 1991, which I find very inspiring. I felt that being part of such an environment would offer a unique learning experience, both academically and personally. In Estonia, you are exposed to a culture of resilience, innovation, and a robust academic system, which I believe shaped my PhD journey in a way that would be difficult to replicate elsewhere.

Why did you decide to pursue a PhD in the first place? 

The main idea and decision started in 2009/2010 at the department of Mathematics when I was a Bachelor student in my university in Syria. At that time, I knew that pursuing a PhD study was an optimal way to learn and discover new research fields, which was for me a kind of passion and curiosity regarding understanding more of our current world through different kinds of sciences, and even potentially have any positive impact. So, I would say that this stems from my innate curiosity to learn more about our world.

What excited you most before you started?

The main PhD topic ‘AI BIAS’ was the most exciting for me before I started.

What alternatives did you seriously consider instead of a PhD? 

I think that I did not have in mind alternatives other than doing a PhD. The goal was very clear for me ‘pursuing PhD’ since my Bachelor studies. I knew that I would do it and finish it someday.

Was there a person, experience, or failure that pushed you in this direction?

There was no person, experience, or failure that pushed me directly in this direction. I think that during master studies, I realized more that I can do that, given that many people pass this PhD journey, so this maybe could indirectly push me to continue in this direction with more motivation.

What did theory teach you about patience, rigour, or intellectual humility? 

Theory taught me patience through the slow process of refining ideas over time. It instilled rigour by requiring every assumption and step to be precise and carefully justified. It also cultivated intellectual humility, reminding me that even simple-looking problems can be far more subtle than they appear, and that my initial understanding is often incomplete and needs to be questioned and refined.

What first exposed you to AI, applied research, or domains like health? 

I think it started during my master’s studies, when I came across posts on social media about the 2017 transformer paper ‘Attention Is All You Need’. That sparked my interest in AI, and from there I began exploring the field more deeply, including taking relevant courses related to machine learning at different universities.

Was this transition from mathematics planned, or did it happen gradually? 

I consider this transition was planned for 2017.

What questions could theory alone not answer? 

Theory can tell us what is possible in principle under ideal conditions, but real-world settings introduce messiness that theory alone cannot fully predict!

What was most intimidating about stepping outside your original field? 

I think I was more intimidated by the idea of staying within my original field than by stepping outside it. I knew there were many exciting areas to explore, such as AI, which motivated me to broaden my perspective and learn beyond my initial domain.

What surprised you most during your first months as a PhD student? 

What surprised me the most in a good way was seeing different new PhD students who had completely different educational backgrounds, such as: neuroscience, physics, philosophy, psychology; and doing PhD in computer science. That was a big motivation for me and made me look like I am not the only one with a different educational background.

What was unexpectedly easier? 

Unexpectedly, turning abstract ideas into concrete experiments became easier than I anticipated. Once the problem was clearly formulated, implementing and testing it often felt more straightforward than the initial theoretical exploration.

When did you first feel the thought: maybe I’m not good enough for this? 

I never really had the feeling that I was not good enough for this path. I had strong trust in the educational environment at the University of Tartu and in the support available throughout my PhD journey, which gave me confidence that consistent effort would lead to progress. In addition, my previous academic experience, my bachelor’s in mathematics in Syria and my master’s studies across several European universities, including the University of L’Aquila, the University of Silesia in Katowice, Vytautas Magnus University, and the University of Turku, also contributed to building this confidence and preparing me for this step.

How did real-world data challenge your theoretical assumptions? 

Real-world data often violated the clean assumptions we relied on in theory, distributions were messy, groups were imbalanced, and hidden correlations created biases we hadn’t anticipated. This forces us to adapt theoretical ideas to handle practical complexities.

When did you truly realise how messy and imperfect data can be? 

I realized it most clearly when evaluating models using some fairness metrics in machine learning. Small changes in the data or group composition led to noticeable shifts in the results. This made it evident that real-world data is often imbalanced, noisy, and sensitive.

What does it feel like to work on problems that affect real people? 

I find working on problems like fairness in machine learning highly motivating and intellectually engaging. At the same time, it brings a strong sense of responsibility, as the outcomes can have real-world consequences. This awareness encourages me to approach my work with greater care and thoughtfulness.

How do you deal with the pressure of wanting to “get it right”? 

I usually focus on clarity and incremental progress rather than perfection. In research, especially on complex problems, ‘getting it right’ is often a process of refinement and repetition, so I rely on careful validation and iteration instead of expecting a perfect answer from the start.

What has been the most difficult moment of your PhD so far?

The most difficult period was at the very beginning of my PhD. Starting in a new country during the COVID-19 pandemic added an extra layer of isolation and uncertainty beyond the usual challenges of a PhD. In addition, I was the first PhD student in my group, which meant developing the research direction largely by myself. While this sometimes meant missing the benefits of a broader research community, it also pushed me to become more independent and self-reliant in my work.

Did you ever seriously consider quitting? Why? 

There were moments, due to various challenges, difficult situations, and external pressures, where quitting felt like a real option. However, I think from the beginning, I was prepared for such obstacles and saw them as part of the journey. I always reminded myself that many others had succeeded in such a journey with less experience or fewer skills, which gave me the confidence to keep going.

What mistake has taught you something important? 

One important mistake was initially trusting theoretical intuition too much without sufficiently validating it empirically. When the results didn’t align, it taught me the importance of combining theory with careful experimentation and being open to revising my assumptions.

When did the interdisciplinary path first feel truly worth it?

It first felt truly worth it when I saw that combining theory with empirical work led to insights I couldn’t have reached from either side alone. In particular, moments where theoretical ideas translated into measurable improvements, or revealed unexpected behavior in practice, made the interdisciplinary approach feel both meaningful and rewarding.

What small success meant more to you than a paper or formal result?

One of the most meaningful small successes was seeing an idea work in practice after some iterations, especially when the results aligned with the intuition I had been developing. I think that that moment of clarity and validation often felt more rewarding than a formal result in a paper.

When did two fields (mathematics and IT) suddenly “click” together? 

It clicked when I saw how theoretical reasoning and empirical machine learning complement each other, especially when theoretical insights helped explain patterns I observed in fairness metrics. That moment, where abstract ideas aligned with real-world behavior, made the connection between the two fields feel natural and powerful.

What part of your work do you quietly enjoy the most?

I most enjoy the moment when an idea I’ve been developing in machine learning is validated and works as intended. It reminds me of my earlier studies, like completing proof of a theorem during an exam, there’s a similar sense of clarity and satisfaction that makes those moments especially rewarding.

What do you enjoy doing when you are not thinking about research?

Outside of research, I enjoy watching science fiction movies and exploring stories that imagine different futures and technologies. I also like reading, spending time outdoors, and having relaxed conversations with friends, which helps me disconnect and recharge.

Do you have any hobbies that help you disconnect? 

Coming from a Mediterranean coastal city, Tartus in Syria, swimming has always been an important part of my life. It helps me clear my mind and fully disconnect, and I find that light physical activity and time away from screens allow me to recharge and return with a fresh perspective.

Is there a place or trip during your PhD that stands out in your memory?

Yes, I attended the 21st Estonian Summer School on Computer and Systems Science in 2024, held at Roosta Holiday Village in northern Estonia. It was a very enriching and memorable experience, both academically and personally. Looking back, the discussions and exposure during that week had a lasting impact on how I think about my research, and after that period I noticed a real shift in my PhD journey, I gained greater clarity and a stronger sense of determination in my research direction.

What do you now appreciate most about living here? 

What I appreciate most about living here is the opportunity to pursue my PhD in such a unique and inspiring environment. If someone had told me ten years ago that I would be doing my PhD in Estonia, I would have been very excited about that possibility. Even though I didn’t fully plan this path, I’m genuinely grateful that it brought me here, and I value the experience and growth I’ve gained from Estonia and all Estonians I have met along the way.

What words or phrases do you use in everyday life?

In everyday life, I naturally use simple Estonian words such as ‘Tere’ for hello and ‘aitäh’ for thank you. Moreover, many of such expressions have become part of my daily vocabulary, and I sometimes find myself using them even outside of Estonia as well.

How has the PhD changed the way you think?

The PhD has changed the way I think by making me more precise, critical, and reflective. I now tend to question assumptions more carefully, break down complex problems into smaller parts, and look for deeper structure rather than surface-level explanations. Through my work on fairness in machine learning, I also became more aware of bias in data, and over time I started noticing similar patterns more clearly in real-world situations as well. Moreover, it has also taught me to be more comfortable with uncertainty and to see it as a natural part of the process of understanding.

How do you feel about the future right now?

After completing my PhD, I feel more positive and confident about most things in the future. The experience has strengthened my motivation to continue working on meaningful problems, especially around fairness and the responsible use of artificial intelligence. I now see my work as part of a longer-term effort to contribute positively to how AI systems are designed and used in practice.

Do you see yourself in academia, industry, or somewhere in between?

I see myself somewhere in between academia and industry. I enjoy the depth and rigor of academic research, but I am also strongly motivated by the practical impact and real-world relevance found in industry.

Which skills from your PhD do you want to carry forward? 

I believe the skills I developed during my PhD have become an integral part of who I am. They have shaped not only my technical abilities but also how I think, approach problems, and handle challenges. For example, critical thinking, rigorous analysis, resilience in the face of setbacks, and the ability to work independently are all skills I will carry forward into the next stage of my journey.

What has moving across theory, AI, and real-world applications taught you about knowledge?

Moving across theory, AI, and real-world applications has taught me that knowledge is not static but continuously refined through interaction between abstraction and practice. Theory provides structure and understanding, while real-world applications reveal limitations, surprises, and new questions that theory alone cannot anticipate. Together, they form a feedback loop where each deepens and reshapes the other.

If your PhD were a chapter in a book, what would its title be?

Equilibrium in Fair and Unfair Learning

Looking back so far—was the uncertainty worth it?

I would say the uncertainty was worth it. While there were many moments of doubt and ambiguity, they were an inevitable part of the process.