BIO

Aleksei Tepljakov received the Ph.D. degree in information and communication technology from the Tallinn University of Technology, in 2015. Since November 2021, he holds a Senior Research Scientist position at the Department of Computer Systems, School of Information Technologies, Tallinn University of Technology. His main research interests include the study of cyber-physical systems: fractional-order modeling and control of complex systems and developing efficient mathematical and 3D modeling methods for virtual and augmented reality for educational and industrial applications with a strong vision to develop intelligent immersive environment. He is a Senior Member of the IEEE with more than 10 years of service. He has been a member of the IEEE Control Systems Society, since 2012, and the Education Society, since 2018.

ETIS

Abstract

In this talk I introduce a novel approach--Muon Beam Imaging (MBI)--a technique with significant potential for non-destructive testing (NDT), industrial and medical diagnostics. While conventional muon tomography offers superior penetration compared to X-rays, its practicality has been limited by the reliance on cosmic rays leading to prohibitively long scan times from hours to weeks and months.

I address this critical limitation by presenting a compact and scalable MBI system enabled by laser-plasma acceleration (LPA). We demonstrate a paradigm shift in muon source technology, leveraging the ultra-high electric fields of LPA to generate a high-flux, artificially produced muon beam. This method not only drastically reduces the physical footprint and power requirements of the system but also provides a tunable source that can be optimized for specific imaging needs, e.g., radiological applications in medicine.

The presentation will detail the methodology (highlighting Machine Learning elements), expected performance metrics, and the multimodal imaging capabilities of this system. It will highlight its broad applicability across key sectors, from industrial quality control in aerospace and nuclear fields to low-dose medical diagnostics. This research represents a crucial step towards making MBI a viable and transformative technology for both science and industry. I also introduce my startup company developing the technology, MuRayTech.

BIO

Dr Andi Hektor is a distinguished particle physicist and the Co-founder and CEO of MuRayTech. He is pioneering the development of Muon Beam Imaging, a very new technology that uses Laser Plasma Accelerators to create artificial muon radiation, mu-rays. This innovative system is poised to revolutionize both industrial quality control, security, and medical diagnostics.

With a background at CERN, University of Cambridge, KBFI in Tallinn,  Andi has spent decades in fundamental science. In 2020 he got a co-founder of GScan, a leader in cosmic-rays-based muon tomography for civil engineering. It is a superb example how to commercialise a high-end technology from fundamental science.  His current venture MuRayTech operating in Tallinn, Hamburg, and Cambridge represents a significant leap forward, moving beyond natural cosmic-ray muons to a compact, high-flux artificial source.

Andi is a dedicated entrepreneur with a proven track record of bridging the gap between academia and industry. He leverages his extensive network and scientific expertise to drive the development of transformative technologies and viable business models for the industrial and medical sectors.

ETIS

Abstract

In this workshop, we develop the semantics of a neural network, step by step, as a certain recursive generator of differentiable functions. This way we learn to see neural networks as just very specific instances to define optimization problems for very high-dimensional functions. Such fundamental viewpoint allows us to break free from the specific two-step weighted-sum/sigmoid-style of neural network definitions. We will analyze Kolmorogorov-Arnold Networks (KANs) in that light. Similarly, the viewpoint even let us break free from gradient-descent as optimization, paving the way for, e.g., quantum annealing for neural network optimization. 

Masayuki Ohzeki, Shuntaro Okada, Masayoshi Terabe, Shinichiro Taguchi (2018). "Optimization of Neural Networks via Finite-Value Quantum Fluctuations". Scientific Reports volume 8, Article number: 9950 https://doi.org/10.1038/s41598-018-28212-4

Shriyank Somvanshi, Syed Aaqib Javed, Md Monzurul Islam, Diwas Pandit, Subasish DasAuthors (2024). "A Survey on Kolmogorov-Arnold Network". ACM Computing Surveys, vol. 58, no. 2, pp. 1-35. https://doi.org/10.1145/3743128

Arun Kumar Sangaiah, Jayakrishnan Anandakrishnan, Sujith Kumar, Gui-Bin Bian, Salman A AlQahtani, Dirk Draheim (2025). "Point-KAN: Leveraging Trustworthy AI for Reliable 3D Point Cloud Completion With Kolmogorov Arnold Networks for 6G-IoT Applications". IEEE Internet of Things Journal. https://doi.org/10.1109/JIOT.2025.3576434

Stephen Chung, Hava Siegelmann (2021). "Turing Completeness of Bounded-Precision Recurrent Neural Networks". Proceedings of NIPS'21 -- the 35th International Conference on Neural Information Processing Systems, Article No.: 2178, Pages 28431-28441. https://dl.acm.org/doi/10.5555/3540261.3542439

Hava Tova Siegelmann, Eduardo D.  Sonta (g1995). "On the Computational Power of Neural Nets". Journal of Computer and System Sciences, vol. 50, pp. 132-150. https://doi.org/10.1006/jcss.1995.1013

BIO

Dirk Draheim is full professor of information systems at Tallinn University of Technology (TalTech) and holds the TalTech National Professorship in Information Society Technology. Dirk holds a Diploma in computer science from Technische Universität Berlin, a PhD from Freie Universität Berlin, and a habilitation from the University of Mannheim. From 2006-2008, he was area manager for database systems at the Software Competence Center Hagenberg, Austria. From 2008-2016 he was head of the data center of the University of Innsbruck and, in parallel, Adjunct Reader at the Faculty of Information Systems of the University of Mannheim. Dirk is co-author of the Springer book “Form-Oriented Analysis” and author of the Springer books “Business Process Technology”, “Semantics of the Probabilistic Typed Lambda Calculus” and “Generalized Jeffrey Conditionalization”. His research interest is the design and implementation of large-scale information systems.

Google Scholar

BIO

From founders and developers to scientists and designers, Gleb Maltsev trains people to speak to audiences with clarity and meaning. Over the course of 14 years, he’s trained thousands of executives to pitch in meetings, events, including trade fairs such as the Mobile World Congress, Smart City Expo, Electronica—and startup founders pitching for capital at Slush, Latitude59, TechBBQ, Startup Day, and TechChill. He regularly works with universities and research institutions—Kaunas, Lappeenranta, Delft University of Technology, TalTech, University of Tartu, University of Latvia.

Most importantly, he’s trained himself. The hard way. By listening to other people's stories, rewriting and delivering them at every workshop, and opening himself up to audience judgment. He’s done all that so others will do the same and grow from it.

ETIS

AI Champions workshop: Practical examples of using ChatGPT as a didactic teaching tool for lecturers

The aim of the workshop is to share practical experiences and solutions on how generative AI (GenAI, ChatGPT) can be used to support teaching and learning at the university. I will introduce practices from my course “Fundamentals of Research Work”, where ChatGPT functions both as a quick end-of-class support tool and as a resource for students’ independent revision and learning. Participants will be introduced to specific didactic techniques, student guidelines, creating in-class quizzes, and designing assignments with the help of GenAI. In addition, I will share experiences from my extended syllabus, including student questions and discussions about the role of GenAI. Participants will gain both new ideas and ready-made practical examples that can be adapted for teaching in their own courses.

Topics

Connection to the DigiComp 3.0 framework: The workshop contributes to developing participants’ skills in “Use and apply AI” and “Evaluate and create AI”, supporting the development of practical abilities to apply AI as well as the capacity to critically evaluate and design AI-based solutions.

Learning outcomes

• Understand how to use GenAI as a short, interactive tool for concept review at the end of a class. 
• Be able to design student guidelines for independent use of GenAI in learning and revision. 
• Gain an overview of different active learning methods that can be combined with GenAI (e.g. pair discussions, panel debates, speed-dating style exams, etc.). 
• Learn to design and adapt quizzes with ChatGPT, and understand the related didactic considerations.
• Be inspired to include GenAI in their course syllabus and to address student feedback and questions about its use.

BIO

Janika Leoste, a PhD in Educational Sciences and a M.Sc. in Economics and Finance, is an Associate Professor of Educational Robotics at Tallinn University and an Assistant Professor of IT Didactics at Tallinn University of Technology. Her research focuses on educational robotics, STEAM education, educational innovation, hybrid and blended learning, AI in teaching and learning, and IT didactics. She has published 60 scholarly articles, including in Frontiers or Francis & Taylor. Since 2019 she is teaching at university, including robot-integrated learning and project based learning projects. She has supervised 12 MA students and has 6 PhD students under supervision. Janika’s previous experience includes developing accounting software and web-based games, producing cartoons and teaching robotics. Janika has successfully applied for research funding from many national and international sources, including the EU’s Horizon Europe – for example, one of her projects is an ERA Talents project EdTech Talents.

ETIS

Abstract

As there is a lot of technology and tools around us, what is and how is trust built towards them? What makes us trust technological tools, and what makes us trust or not trust AI? The lecture focuses on both humans natural instincts, risk awareness and the cost of risk of trusting the environment and machines around us. In the lecture also natural intuition is challenged using numeric methods as a trust building mechanism in the explainable AI context.

BIO

Mr. Jaak Kapten is a software, data science and machine learning engineer with 25+ years of private sector experience. Key specialization areas have been civil engineering quality assurance equipment, software and processes; specialized data centric analytical software design and machine learning solutions of investment risk assessment for capital markets applications. In Taltech Jaak Kapten gives lectures as assistant lecturer in Data Mining, Machine Learning and Foundations of Data Science courses in the Department of Software Science. Current field of research in Taltech is machine learning and explainable AI in the context of resilience, as part of the Next Gen Digital State Research Group.

LinkedIn

Bio

Joosep Pata is a senior research scientist working on computing and AI applications for CERN. He holds a PhD in physics from ETH Zurich and did a postdoc at Caltech. His research focuses on applied and high-performance AI models for challenging data reconstruction and analysis tasks involving real-world experimental data.

ETIS

Session Introduction

This hands-on session focuses on using Natural Language Processing (NLP) and AI-based topic modeling to analyze textual data. Participants will work with techniques like Latent Dirichlet Allocation (LDA) to identify patterns in text datasets such as academic articles, customer reviews, or social media posts.

The session covers the research workflow from data collection to visualization. Dr. Aleem will demonstrate methods used in published studies on remote work, leadership trends, and entrepreneurship. Participants will use Python-based tools to see how these techniques work in practice.

This session is for researchers, doctoral students, and practitioners who want to analyze text data. Basic familiarity with research methods is helpful, but no programming experience is required.

Practical Workshop: Applying Topic Modeling to Your Data

Participants will work through the topic modeling process:

• Define a research question: Select a text analysis problem from your field (customer feedback analysis, literature reviews, policy documents, online discussions).
• Prepare data: Clean text, remove noise, and structure data for modeling. Learn which preprocessing steps matter and which don't.
• Run LDA model: Apply LDA to sample datasets. 
• Interpret outputs: Read topic distributions, identify meaningful themes, and validate results. Use visualization tools like pyLDAvis and word clouds.
• Troubleshoot common issues: Address problems like overlapping topics, and choosing the right number of topics.

Learning Outcomes

By the end of this session, participants will:

• Understand how topic modeling works and where it can be applied in research.
• Know the basic steps in a topic modeling workflow, from data collection to interpretation.
• Be able to read and interpret topic model outputs.
• Understand common challenges in text analysis and how to address them.

BIO

Majid Aleem is a researcher and University Lecturer in International Business at the Turku School of Economics, University of Turku, Finland. His research focuses on remote work, digital collaboration, and how AI affects people management and organizational behavior.

His work has been published in journals including the Journal of Business Research, Technological Forecasting & Social Change, and Journal of Retailing and Consumer Services. He uses computational methods like topic modeling and machine learning to study workplace changes. Recent projects examine AI's impact on leadership in remote work and human resource management.

Majid has taught in Finland, the UK, Sweden, Estonia, Russia, and Pakistan. He co-developed GVTanalytics, a platform for teaching team collaboration, and created a GPT assistant for his International Business Strategy course. He trains doctoral researchers and faculty on using AI in research and teaching, and works as AI Oversight Manager for the ARCANA Eurostars project. He has supervised 43 master's theses and 6 bachelor's theses.

Selected Publications

• Aleem, M., Sufyan, M., Ameer, I., & Mustak, M. (2023). Remote Work and the COVID-19 Pandemic: An Artificial Intelligence-Based Topic Modeling and a Future Agenda. Journal of Business Research, 154, 113303.
• Hossain, M., Ahmed, F., Aleem, M., Talaoui, Y., Bask, A., & Rajahonka, A. (2025). Emerging technologies in sharing economy: A review and research agenda. Technological Forecasting and Social Change, 211, 123885.
• Mustak, M., Hallikainen, H., Laukkanen, T., Ple, L., Hollebeek, D.L., & Aleem, M. (2024). Using machine learning to develop customer insights from user-generated content. Journal of Retailing and Consumer Services, 81, 103995.

ETIS

Session Introduction

This interactive session introduces Global Team Coach (GTC), a pioneering Intelligent Tutoring System (ITS) designed to personalize learning in intercultural, team-based education using Generative AI. Developed and implemented in a Master’s-level international business course, GTC responds dynamically to students' emotional, behavioral, and cognitive cues during global virtual teamwork. The system supports metacognitive reflection, leadership development, and cultural intelligence through situated, just-in-time feedback. Participants will explore the system’s adaptive architecture, including its learner model, content framework, and instructional strategies built around real-world data from over 1,000 student reflections. Drawing on adaptive learning theory and GenAI capabilities, this session will demonstrate how intelligent systems can offer empathic, developmentally attuned coaching in higher education settings. Participants will leave with insights into the challenges and possibilities of integrating GenAI into ITS design, especially in complex, socially rich learning environments.

Creative Workshop: Enhancing ITS in Higher Education
Participants will engage in a hands-on creative workshop, using design thinking methods, to conceptualize their own intelligent tutoring system tailored for their learning contexts. Working in small groups, they will:

• Define a learning challenge in their own discipline (e.g., teamwork, leadership, ethics, communication).
• Draft an adaptive learner model using traits, needs, and cues relevant to their student populations.
• Design micro-interventions or feedback strategies based on GenAI capabilities (e.g., reflective prompts, emotional resonance, peer modeling).
• Discuss ethical, pedagogical, and technical considerations in deploying GenAI tools in education.
  Each group will present their concept to receive peer feedback, sparking dialogue about scalable innovation and responsible AI use in higher education.

Learning Outcomes

By the end of this session, participants will be able to:

• Understand how GenAI can enhance personalized and adaptive learning in complex team-based environments.
  Identify key components of an ITS, including dynamic learner models and instructional scaffolding logic.
• Design context-specific enhancements to existing or future ITSs aligned with affective and cognitive learner development.
• Reflect critically on the ethical and pedagogical implications of integrating GenAI into higher education.

Bio

Peter is Associate Professor and University Research Fellow in International Business at the Turku School of Economics, University of Turku, Finland. With over two decades of experience in academic teaching, research, and executive education across Europe, Oceania, and Latin America, he is recognized for his contributions to experiential and adaptive learning in international business education. Peter co-founded the Global Innovation Management (GIM) program and has supervised numerous award-winning doctoral theses. His current research explores the intersection of artificial intelligence, global virtual teamwork, and personalized learning. He is co-developer of Global Team Coach, a GenAI-powered Intelligent Tutoring System designed to support intercultural collaboration and leadership development in diverse student teams. As a frequent contributor to international conferences and interdisciplinary projects, Peter is committed to bridging theory and practice, helping learners and educators navigate the complexities of global work and digital transformation.

Google Scholar

BIO

Research Professor at Tallinn University of Technology, specializing in artificial intelligence in education and intelligent assessment methods, and an Assistant Professor at the University of Novi Sad. His research spans AI-driven educational solutions, digital assessment systems, service engineering, and digital servitization. In recent years, he has focused on the responsible and ethical integration of AI in learning environments, contributing to several Horizon Europe and Erasmus+ projects on educational technology. Dr. Rakić serves as a National Ambassador for the European Commission’s DigiEduHack initiative and was a stakeholder expert in the development of DigComp 3.0, advising on AI-related competences. He has also acted as an industrial development expert for UNIDO, bringing global perspective to policy and innovation. His academic record includes 34 publications indexed in Scopus with an h-index of 14. This combination of scientific expertise, EU policy engagement, and global advisory roles positions him as a key voice on AI ethics and educational innovation.

ETIS

BIO

Tarmo Lipping got his DrTech degree in Signal Processing in 2001 and MBA in 2013 from Tampere University of Technology. In 2001-2002 he was postdoctoral research associate at Dartmouth College, NH, USA. In 2002-2003 he was director of the Biomedical Engineering Center and held professorship in Biomedical Engineering in Tallinn University of Technology. Since 2004 he is professor of Signal Processing in Tampere University. During 2019-2023 he served as the director of Pori University Consortium. His research interests include applying machine learning and artificial intelligence tools to the monitoring of cognitive and mental states in real-life situations. During his career, Tarmo Lipping has been Principal Investigator of numerous academic and industry-related research projects; he is author or co-author of over 100 research publications and has supervised over 60 Masters' and 7 Doctoral theses.

Google Scholar