News and events archive

In November, Šárka Pospíšilová, Vice-Rector for Research and Doctoral Studies at MU, honored newly graduated doctoral students for their excellent academic results. Two teams received awards on behalf of the Faculty of Informatics: Agáta Kružíková and her supervisor Václav Matyáš, and Michal Štefánik and his supervisor Petr Sojka. What advice would they give to students who are hesitant about applying for a doctoral program? Where did the graduates find employment? And what qualities distinguish students who are able to write an exceptional thesis? Our award winners answer these and other questions:

RNDr. Agáta Kružíková, Ph.D.

Agáta Kružíková graduated from the Faculty of Informatics with a degree in Informatics, specializing in the Fundamentals of Informatics. During her studies, she focused on improving the usability and security of user authentication in computer systems. Her research expanded theoretical knowledge and led to practical recommendations that have already influenced the behavior of IT professionals and solutions in mobile banking.

What does this award mean to you personally?

I see the award as significant recognition of the several years of work and dedication I have devoted to research. At the same time, it is also an appreciation of the support and guidance of my supervisor and confirmation that we have devoted ourselves to a current and socially and scientifically relevant topic.

What would you recommend to students who are concerned that a doctorate is "detached from reality" and will not be useful in practice?

I recommend establishing cooperation with industrial partners during your studies and working together to find a topic that will be beneficial for both sides. Dialogue with people from the field makes it easier to identify areas for improvement. Finding solutions to these areas is more easily transferable to practice for the relevant partner, thereby increasing the impact of the entire doctoral thesis.

Our collaboration with industry partners took place in this way—we held joint discussions in which we tried to identify current challenges from practice that also offered scientific benefits. Based on these discussions, we designed a user study focused on the perception of the usability and security of authentication methods in mobile banking. From the results, we then formulated specific recommendations, some of which were implemented in a mobile banking application used in the Czech Republic.

Where are you headed after graduation?

After several years in academia, I moved to the private sector, where I can put my knowledge of cybersecurity into practice—I work as a cybersecurity consultant at PwC.

Prof. RNDr. Václav Matyáš, M.Sc., Ph.D.

Václav Matyáš is a professor of computer science and head of the Center for Research in Cryptography and Security at the Faculty of Informatics, Masaryk University. His areas of expertise include applied cryptography, cybersecurity, information privacy protection, and academic-industrial relations at the national and international levels.

In your opinion, what characteristics distinguish students who are able to write exceptional final theses?

Not only are they prepared, but they usually go the extra mile and do extra work on their own, resolve any minor shortcomings, and do not hesitate to devote time to promoting their research results, etc. Simply put, they go to meet their luck, because recognition is often a matter of luck when preparedness and willingness to look at things from different perspectives count as luck.

What surprised or impressed you most about working with Agáta Kružíková?

I was pleased by her commitment and willingness to communicate not only with all research participants, but often also with those who were yet to be involved in the research. I was also glad that she believed in me in the early stages of the study, when the research took longer than in many other areas of computer science, and publications and other indicators of success came later.

I was surprised by comments from some corners of the faculty that this research was "not sufficiently computer science-oriented." But on the other hand, I had already heard these comments about another PhD student, so I got used to them.

How do you motivate students when they feel that their research is not progressing?

I share with them both good and bad experiences—my own and those of other members of our research center. And if they are interested, I allow them to participate partially in another research project to diversify and gain a different perspective. But I am still learning in this area and consider my skills to be insufficient.

Mgr. Michal Štefánik, Ph.D.

Michal Štefánik graduated from the Faculty of Informatics with a degree in Informatics, specializing in the Fundamentals of Informatics. He focused on the error rate of commonly used artificial intelligence language models. He proposed specific procedures to improve the resilience of models to unexpected inputs and thus the reliability of results. The new findings are used to develop safer and more robust AI systems.

What does this award mean to you personally?

I see the award as recognition of my approach to doctoral studies, which was in many ways non-standard in my field, and I believe it can also be an inspiration to others.

First and foremost, I decided to focus exclusively on the most prestigious forums in my field, which I was able to attend regularly thanks to the thoughtful and well-organised support at FI. This allowed me to establish close contacts with the "first league" of research in the field of NLP and AI and to understand which topics deserve attention. Participating in top conferences on other continents was challenging, but in retrospect, I find it very rewarding on a personal level. I realized that problems can be solved in completely different ways, and I understood the barriers of my social bubble.

Secondly, in my studies, I focused heavily on supporting younger students who were interested in and enjoyed scientific work. During my studies, I supervised 10 theses, which resulted in 7 awards, including 3 international ones, and later 3 publications at top CORE A/A* conferences. Since 2022, I have organized regular meetings of the TransformersClub student club, which continue even after my departure from FI.

It may seem that supporting students is just a voluntary activity, but it has allowed me to learn a lot, such as understanding technologies that I would never have encountered otherwise. The main challenge of good cooperation is to organize meetings with students in a fully open atmosphere and build on internal motivation.

What would you recommend to students who are afraid that a doctorate is "detached from reality" and will not be useful in practice?

A doctorate in itself means nothing – doctoral studies are what you make of them. I worked part-time throughout my studies, which gave me a lot of inspiration for my research, especially at the beginning. Later, as part of my studies, I also helped with the design and implementation of a commercialized project between the faculty and an industrial partner. I think it's not difficult to stay close to real-world practice during doctoral studies in computer science. However, it's good to identify with the role of "architect" of your own path and look for creative solutions, such as getting involved in communities that open up opportunities for collaboration with the real world.

Where are you headed after completing your studies?

Although I certainly did not expect it at the beginning of my studies, over time I realized that the freedom to choose my own goals, which research and academia provide, is important to me. Academia also continues to provide me with opportunities for personal and professional growth.

After completing my studies, I accepted a postdoctoral position at the University of Helsinki, and recently decided to continue as a researcher at the newly established R&D Center for Development of Large Language Models (LLMC) at the National Institute of Informatics in Japan. The LLMC is a strategic investment funded directly by the Japanese government, which means that, in addition to our status as government employees, we have access to thousands of Nvidia H200 GPUs and interesting personalities in the field. For example, I work under the supervision of Prof. Pontus Stenetorp, who is also a full professor at University College London beyond the LLMC.

The main goal of my research is to understand and improve the robustness of language model decision-making. In our project, together with mathematicians from Oxford, we focus on evaluating and understanding the best mathematical models in the AIMO competition.

Assoc. Prof. RNDr. Petr Sojka, Ph.D.  

Petr Sojka works at the Department of Visual Informatics, Faculty of Informatics, Masaryk University. His research focuses on electronic publishing, digital mathematical libraries, digitization of mathematics and information retrieval, gait recognition, and natural language processing. In particular, he deals with language issues in high-quality typesetting, engineering of electronic publishing and digitization projects, and machine learning. He leads the Laboratory of Electronic Multimedia Applications (LEMMA) at FI MU.

In your opinion, what characteristics distinguish students who are able to write exceptional final theses?

Internal motivation, perseverance, the ability to concentrate, and leaving their comfort zone. In the words of Edison:  "Genius is one percent inspiration and ninety-nine percent perspiration" or Einstein: "It's not that I'm so smart, it's just that I stay with problems longer."

Then there's the ability to communicate and a growth mindset – actively seeking feedback, enjoying your mistakes, and learning from them. A poem by Piet Hein, which Donald Knuth reminds himself of every day when he returns home:

The road to wisdom? Well, it's plain
And simple to express:
Err
and err
and err again,
but less
and less
and less.

What surprised or impressed you most about working with Michal Štefánik?

In the nearly decade since spring 2016, when Michal enrolled in my Introduction to Information Retrieval course, he has made incredible progress in terms of his personality, knowledge, and scientific abilities. It was a joy to watch him grow from a practical and pragmatic programmer who had to retake his state exams twice, from publication to publication, gaining confidence, developing communication and presentation skills, gradually becoming independent, and growing into an independent scientist graduating summa cum laude and building his own research group.

And now, at a prestigious workplace in Tokyo, he is writing grants for his next research project during his second postdoc. The turning point came with his first publications as first author and conference successes, which he enjoyed and found motivating.

How do you motivate students when they feel that their research is not progressing?

Every student is an individual and each one is unique, so the key is to figure out whether the stagnation is due to a lack of internal motivation, creativity, self-confidence, or an inability to concentrate. I try and learn to motivate my students in a positive way, supporting them in a good way, giving them confidence, and anticipating their successes.

I think about their talents, praise them for them, and support them in identifying and developing them so that their successes trigger a spiral of growth in internal motivation, self-confidence, and independence.

It is also effective to guide them in applying Feynman's principle, learning and research techniques expressed by the challenge "If you can't explain it simply, you don't understand it well enough" and the quote "The first principle is that you must not fool yourself — and you are the easiest person to fool."

We congratulate all the award winners and believe that their example will motivate other students who are considering doctoral studies.

Author: Marta Vrlová, Office for External Relations and Partnerships at FI MU

Photos: Martin Indruch, Agáta Kružíková's archive, Michal Štefánik's archive