Physician / ML engineer / PhD candidate
Applied scientist, ML engineer, physician, and PhD candidate building real-world AI across medicine, multimodal models, and deployment-minded systems.
18
Papers and abstracts
9
Conference presentations
PhD
Renal outcomes focus
6
Core collaborators
Practice
Work mode
Research design, predictive modelling, evaluation, and deployment-minded systems work carried end to end.
Core domains
Medicine, NLP, computer vision, multimodal learning, world models, registry-scale analysis, and technical infrastructure for ambitious applied work.
Operating standard
Methods should stay rigorous, interpretable, and useful once they meet real data, real workflows, and implementation constraints.
Focus areas
Doctoral work on fairness, transparency, and clinically usable evaluation in nephrology and acute care.
Now published in Journal of Clinical Medicine as part of the doctoral renal prediction line.
Read publicationProject Quintessence
Independent foundation work connecting clinical AI, open systems, and technically ambitious collaborative builds.
Visit websiteRegistry-scale work
Ongoing cohort design and translational modelling on institutional-scale clinical data.
Work mode
Research design, predictive modelling, evaluation, and deployment-minded systems work carried end to end.
Core domains
Medicine, NLP, computer vision, multimodal learning, world models, registry-scale analysis, and technical infrastructure for ambitious applied work.
Operating standard
Methods should stay rigorous, interpretable, and useful once they meet real data, real workflows, and implementation constraints.
Clinical prediction
Doctoral work on fairness, transparency, and clinically usable evaluation in nephrology and acute care.
Multimodal systems
Evaluation-focused work on multimodal models, alignment, and deployment-minded reliability.
Clinical AI
Extraction systems, chest X-ray workflows, and tool-using patterns built with auditability in mind.
Infrastructure
Large clinical cohorts, reproducible data systems, and collaborative infrastructure through Project Quintessence.
Now published in Journal of Clinical Medicine as part of the doctoral renal prediction line.
Read publicationProject Quintessence
Independent foundation work connecting clinical AI, open systems, and technically ambitious collaborative builds.
Visit websiteRegistry-scale work
Ongoing cohort design and translational modelling on institutional-scale clinical data.
Featured work
AKI
Journal of Clinical Medicine / 2026Forecasting ICU Acute Kidney Injury with Actionable Lead Time Using Interpretable Machine Learning: Development and Multi-Center Validation
Doctoral anchor
Extends kidney injury prediction toward real lead-time utility instead of retrospective signal mining alone.
NPI
BMC Infectious Diseases / 2025The Impact of COVID-19 Non-Pharmaceutical Interventions on Notifiable Infectious Diseases in Poland: A Comprehensive Analysis from 2014-2022
Population health analysis
Connects model thinking to surveillance, policy shifts, and long-horizon epidemiologic interpretation.
Research
AKI
Journal of Clinical Medicine / 2026Forecasting ICU Acute Kidney Injury with Actionable Lead Time Using Interpretable Machine Learning: Development and Multi-Center Validation
Doctoral anchor
Extends kidney injury prediction toward real lead-time utility instead of retrospective signal mining alone.
NPI
BMC Infectious Diseases / 2025The Impact of COVID-19 Non-Pharmaceutical Interventions on Notifiable Infectious Diseases in Poland: A Comprehensive Analysis from 2014-2022
Population health analysis
Connects model thinking to surveillance, policy shifts, and long-horizon epidemiologic interpretation.
Writing
Research Notes
Jun 17, 2026The Patient Is Not a Label: Toward World Models of Physiology
26 min read
Clinical AI has spent a decade scoring labels. The next frontier is action-aware patient-state modeling: physiology, measurement, intervention, uncertainty, and consequence.
Clinical AI
Jan 31, 2026Clinical Note Extraction in the Age of Generative AI: Error Patterns, Methodological Architectures, and Strategic Implementation
22 min read
An in-depth survey of clinical note extraction methods--rule-based, BERT, and LLMs--with error patterns, taxonomy of failures, and implementation strategy.
Collaborators
Medical University of Warsaw
Khalifa University
European Renal Association
Quintessence Health AI
SimplePod
Voltage Park
Contact