JAKUB MACINA
HUMAN-CENTRIC AI (FOR) LEARNING
AI & ML Researcher, Engineer, Educator, Entrepreneur and Speaker.
PhD at ETH Zurich, machine learning (ML) and natural language processing (NLP) for more effective human-AI collaboration.
Experience
Experience in software engineering, machine learning research, education and e-commerce. Jakub is Forbes 30 under 30 in the category of Science and Education.
Research-oriented approach: with publications in top-tier conferences such as EMNLP, ACL, RecSys, and NeurIPS. Received ETH AI Center Fellowship (<4% acceptance rate) and research grants in value over 100 000 CHF. Advised by Prof. Mrinmaya Sachan and Prof. Manu Kapur.
Applied research: as a machine learning engineer and senior staff machine learning engineer in one of the fastest-growing startups in Europe (which was acquired by Bloomreach in 2021). Responsible for building large-scale machine learning training and inference pipelines from scratch for a product recommender system platform used by hundreds of clients.
Teaching: I have taught courses on machine learning, large language models, and natural language processing in education at ETH Zurich. I have also given talks at various conferences and meetups.
Leadership: Both at the university and professionally, I successfully managed a team of six data scientists and advised more than 12 Master's and Bachelor's theses at ETH Zurich.
Entrepreneurship and Product management: I have led projects and defined product requirements that have been successfully launched at scale. Co-founded a startup in a health-tech space with a seed investment, was a member of Entprepreneur First cohort in Berlin.
Large-scale infrastructure architecture: experience designing and implementing training and inference pipelines at scale, with the ability to handle thousands of requests per second using Kubernetes in Google Cloud Platform.
Open-source development: I have contributed to open-source projects such as Discourse discussion platform (also with Google Summer of Code), and I have developed a few projects myself.
Research
Artificial Intelligence, Machine Learning, Natural Language Processing, all combined with Human Computer Interaction.
Understanding the reasoning and pedagogical capabilities of generative large language models (LLMs such as GPT, T5, Flan-T5, BART, LLAMA).
Neural Language Generation and Dialogue Systems, focus on Pedagogical Alignment and Fine-tuning of LLMs for Tutoring.
Reasoning and Math Word Problem Solving.
Evaluation of Large Language Models.
Question Answering and Question Generation.
Recommender Systems and Personalization.
Talks
2024 | Battle: Human vs. AI, Forbes Business Fest, Bratislava |
2023 | LLMs for Tutoring: Socratic Questioning and Scaffolding, AI in Education, AI+X Summit 2023, Zurich |
2023 | Talkshow about AI, Podcast by Forbes Slovakia |
2022 | Feedback Generation for Personalized Learning, ETH AI Center Doctoral Symposium, ETH Zurich |
2019 | Deep Learning for Recommender Systems, Vienna Deep Learning Meetup, Vienna |
2018 | Realtime Personalized Customer Embeddings, Python Conference (PyCon), Bratislava |
2018 | Tuning Search at Scale (Discourse open source project), Google Developer Group (GDG), Zurich |
2018 | Introduction to Natural Language Processing, Data Science Club by Exponea, Bratislava |
2018 | Embedding-based Product Recommendation, Google Launchpad, Warsaw |
Teaching
2024 | Artificial Intelligence in Education, Head TA, ETH Zurich |
2024 | Large Language Models, ETH Zurich - Lecture Notes |
2022 | Artificial Intelligence in Education, Head TA, ETH Zurich |
2021 | Artificial Intelligence in Education, ETH Zurich |
2019 | Text-based Information Retrieval, Slovak University of Technology |
2019 | Artificial Intelligence, Slovak University of Technology |
2018 | Text-based Information Retrieval, Slovak University of Technology |
Publications
Romain Puech, Jakub Macina, Julia Chatain, Mrinmaya Sachan, Manu Kapur
preprint
Nico Daheim*, Jakub Macina*, Manu Kapur, Iryna Gurevych, Mrinmaya Sachan
EMNLP 2024
Video
Junling Wang, Jakub Macina, Nico Daheim, Sankalan Pal Chowdhury, Mrinmaya Sachan
ACL 2024 Findings
Jakub Macina*, Nico Daheim*, Sankalan Pal Chowdhury*, Tanmay Sinha, Manu Kapur, Iryna Gurevych, Mrinmaya Sachan
EMNLP 2023 Findings
Video
Jakub Macina*, Nico Daheim*, Lingzhi Wang, Tanmay Sinha, Manu Kapur, Iryna Gurevych, Mrinmaya Sachan
EACL 2023
Video
Kumar Shridhar*, Jakub Macina*, Mennatallah El-Assady, Tanmay Sinha, Manu Kapur, Mrinmaya Sachan
NeurIPS 2022, 2nd MATH-AI Workshop
Video
Kumar Shridhar*, Jakub Macina*, Mennatallah El-Assady, Tanmay Sinha, Manu Kapur, Mrinmaya Sachan
EMNLP 2022
Video
Michal Kompan, Peter Gaspar, Jakub Macina, Matus Cimerman, Maria Bielikova
IEEE Intelligent Systems ( Volume: 37, Issue: 1, 01 Jan.-Feb. 2022)
Jan Paralic, Tomas Kaszoni, Jakub Macina
Advances in Intelligent Systems and Computing 2019
Peter Kovacs, Matus Cimerman, Jakub Macina
12th Workshop on Intelligent and Knowledge-oriented Technologies 2017
Jakub Macina, Ivan Srba, Joseph Jay Williams, Maria Bielikova
RecSys 2017: Proceedings of the Eleventh ACM Conference on Recommender Systems