Jakub Binda

Developing ML solutions for construction industry.

Conducting research at the intersection of Computer Vision and Explainable AI (XAI) in the medical domain, with a focus on interpretability of Vision Transformers.

I was involved in the international INFORM consortium project (https://inform-project.eu/) which aimed to develop explainable deep neural network models for the medical image analysis. I was responsible for training deep neural network models based on CNN, ViT architectures on medical images and applying existing or developing novel post-hoc explainability tools. Finally, I pursued an Advanced Master Studies program at University of Warsaw, where I was involved in a research project focused on interpretable machine learning models for head and neck cancer recurrence prognosis (survival analysis).

Led the end-to-end development of Obz.AI (https://obz.ai/), an MVP platform for monitoring vision-based AI models in production, including designing and publishing a Python package, architecting a FastAPI backend with a database, and coordinating frontend development of a React web application.

Developed and led the creation of a Python package providing post-hoc model explainability and anomaly detection tools for the image domain.

I joined research group held by Prof. Jacek Jemielity and Dr Joanna Kowalska. There I was taking part in the research project focused on developing fluorescently labelled RNA probes for decapping studies. In particular, I was involved in the organic synthesis of fluorescently labelled trinucleotide cap analogs. Most important techniques which I have used: - Organic synthesis in solvents and on solid phase. - Reversed Phase High Performance Liquid Chromatography. - Ion Exchange Liquid Chromatography. - Thin Layer Chromatography. - UV-Vis Spectroscopy. - Mass Spectrometry. - Nuclear Magnetic Resonance Spectroscopy.

I supported the development and operational activities of a MedTech startup. My responsibilities included assisting in project management, coordinating tasks with project management tools (e.g., ClickUp), preparing and reviewing contractual documentation with subcontractors and identifying potential collaborators and suppliers. I also contributed to preparing documentation for certification bodies (e.g., DQS, TÜV NORD) and gained practical exposure to the medical device production process - from component sourcing to final sterilization.

Project aimed to cultivate and sell carnivorous plants. Among the activities worth emphasizing are: obtaining the right to passport plants, activities promoting the offer, organizing online sales and stationary sales in the form of an exhibition stand. Project fanpage: Facebook: https://www.facebook.com/FlorigenRosliny/

During studies I specialized in machine learning and deep learning applied in medicine.

In order to gain an interdisciplinary foundation, I combined biology/chemistry subjects with mainly mathematics and data analysis subjects. I graduated with honors. Thesis topic: Creation of trinucleotide tools for studying RNA decapping.

Polish Children's Fund is an organisation that supports talented children and teenagers in Poland. I took part in programs organised by this institution. For example, I attended biological workshops in Warsaw.