PhD Researcher | Chief Technology Officer | AI Research Scientist
Technical University of Denmark | University of Luxembourg | Skolyn

I am a PhD researcher and AI scientist working at the intersection of human-AI collaboration, medical imaging, and systems biomedicine. Currently pursuing dual doctoral degrees at DTU Compute (Human-XAI Collaboration for Fetal Ultrasound Imaging) and University of Luxembourg (Systems and Molecular Biomedicine), while serving as CTO and Co-Founder of Skolyn, a clinical AI company developing explainable diagnostic systems.

My research spans medical AI, computational biology, federated learning, urban intelligence, and trustworthy machine learning. I have contributed to projects at Google Health AI, Finnish Center for Artificial Intelligence, Uppsala University, European Research Council, and multiple leading institutions across Europe and beyond. My work emphasizes explainability, privacy preservation, and human-centered AI design with real-world clinical validation.

PhD Researcher - Technical University of Denmark, DTU Compute (Feb 2025 - Mar 2028)
Human-XAI Collaboration for Fetal Ultrasound Imaging | GPA: 4.0/4.0 | Research Excellence Award '25

PhD Researcher - University of Luxembourg (Feb 2025 - Jan 2028)
Systems and Molecular Biomedicine | GPA: 4.0/4.0 | University Merit Scholarship '25

Postdoctoral Researcher - Uppsala University, Vi3 Division (Jul 2025 - Present)
Medical AI & Imaging, Neuroradiology, Federated Learning | 3+ image processing pipelines, 95%+ AUC diagnostics

Data Science Specialist in Proteomics - DTU Bioengineering (Jan 2025 - Present)
5 bioinformatics pipelines processing 10+ TB data annually | 30% computation runtime reduction

Peer Reviewer - IEEE Journal of Biomedical and Health Informatics (Feb 2026 - Present)
Scientific novelty, methodological rigor, reproducibility, biomedical relevance

Chief Technology Officer & Co-Founder - Skolyn (Jan 2025 - Present)
Clinical Co-Pilot AI Platform | 127 pathological indicators in <3s | 50,000+ scans processed | CE Mark, ISO 13485, FDA 510(k) | $2M Seed round

Senior AI Engineer - WUF13 Azerbaijan (Jan 2026 - Present)
13th World Urban Forum AI & Analytics | 13,000 participants, 166 countries | Forecasting, anomaly detection, computer vision

Senior AI Engineer - State Committee on Urban Planning and Architecture, Azerbaijan (Jan 2026 - Present)
Geospatial data engineering, ML for planning/zoning, governance-aligned deployment

Fachberater für Künstliche Intelligenz - Bundeswehr (German Federal Armed Forces) (Dec 2025 - Feb 2026)
Military Language Models, Strategic AI Development, NATO Compliance, NLP Frameworks

Kaggle Benchmarks Trusted Tester - Kaggle Early Access Program (Oct 2025 - Feb 2026)
Task-based evaluations, benchmark validation, SDK testing, safety assessment

Seconded National Expert, Policy Analysis - European Research Council (ERC) (Nov 2025 - Dec 2025)
Horizon Europe research programmes (16B euros) | 12,000+ ERC projects, 75,000 researchers, 200,000+ publications

Research Scientist - Generative AI Evaluations - Google Health AI (Oct 2024 - Dec 2024)
50 board-certified clinicians, 1M+ patient interactions | 18% bias reduction | HIPAA, GDPR, ISO 13485

Technical Program Manager II - Google Health (Apr 2024 - Sep 2024)
Federated Health Data Platform | 7 clinical research partners | 99.97% uptime

Senior Research Scientist - Finnish Center for Artificial Intelligence (Jul 2024 - Aug 2024)
Multi-Omics Clinical Intelligence Platform | 50,000+ patient records | 2 publications in Nature Medicine and Patterns

Associate Research Scientist - FCAI (Oct 2023 - Jun 2024)
Adaptive Federated Ensembles | AUC 0.91→0.95 | 2 publications (IEEE TMI, NeurIPS XAI Workshop)

Research Scientist - FCAI (Jan 2023 - Sep 2023)
ARL-FCAI-03 RL framework | 1.5M MRI slices | 22% error reduction, 30% AI-human agreement improvement

Bioinformatician to Clinical Genomics - Linköping University (Jul 2025 - Dec 2025)
5+ NGS pipelines (Nextflow/nf-core) | 2TB/week genomic data | 500+ clinical samples annually

Adjunct Instructor - Linköping University (Jan 2025 - Nov 2025)
6 courses (Statistics, Machine Learning, Python) | 150+ students/semester | 90%+ satisfaction | Teaching Award nomination

Doctor of Philosophy (PhD) - Technical University of Denmark (Feb 2025 - Mar 2028)
Human-XAI Collaboration for Improved Fetal Ultrasound Imaging
GPA: 4.0/4.0 | Research Excellence Award '25 | Dean's List

Doctor of Philosophy (PhD) - University of Luxembourg (Feb 2025 - Jan 2028)
Systems and Molecular Biomedicine
GPA: 4.0/4.0 | University Merit Scholarship '25 | Dean's List

Master of Science (MS) - Linköping University (Aug 2024 - Feb 2026)
Statistics and Machine Learning
GPA: 3.95/4.0 | Excellence Scholarship '24 | Dean's List

Bachelor of Science (BS) - Tampere University (Aug 2021 - Jun 2024)
Computing and Electrical Engineering
GPA: 3.9/4.0 | President's Medal '24 | Dean's List | Merit Scholar

International Baccalaureate Diploma - International School of Helsinki (Jul 2019 - Jun 2021)
Score: 40/45 (top 6% cohort) | Academic Excellence in STEM '21

Urban Spatio-Temporal Foundation Models for Climate-Resilient Housing: Scaling Diffusion Transformers for Disaster Risk Prediction
arXiv, Feb 2026 | Skjold-DiT framework, Baltic-Caspian Urban Resilience dataset (847,392 buildings)

Autonomous AI Agents for Real-Time Affordable Housing Site Selection: Multi-Objective Reinforcement Learning Under Regulatory Constraints
arXiv, Feb 2026 | AURA system, 94.3% regulatory compliance, 37.2% Pareto improvement

Energy-Efficient Neuromorphic Computing for Edge AI: A Comprehensive Framework with Adaptive Spiking Neural Networks and Hardware-Aware Optimization
arXiv, Feb 2026 | NeuEdge framework, 847 GOp/s/W efficiency, 67% energy reduction

Multi-Fidelity Physics-Informed Neural Networks with Bayesian Uncertainty Quantification and Adaptive Residual Learning
arXiv, Feb 2026 | MF-BPINN framework, 73-86% computational cost reduction

PatchFormer: A Patch-Based Time Series Foundation Model with Hierarchical Masked Reconstruction and Cross-Domain Transfer Learning
arXiv, Jan 2026 | 24 benchmark datasets, 27.3% MSE reduction, 94% less training data

Transparency-First Medical Language Models: Datasheets, Model Cards, and End-to-End Data Provenance for Clinical NLP
arXiv, Jan 2026 | TeMLM artifacts, 498,000 synthetic clinical notes, ProtactiniumBERT

Hierarchical Attention-Enhanced Graph Neural Networks with Deep Reinforcement Learning for Large-Scale Stochastic Vehicle Routing Problems
arXiv, Jan 2026 | HAT-GNN-RL, 500+ customer instances, 0.8-2.3% optimality gap

Mechanistic Analysis of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning
arXiv, Jan 2026 | 109B-400B parameters, gradient interference analysis

Strategic Doctrine Language Models: A Learning-System Framework for Doctrinal Consistency and Geopolitical Forecasting
arXiv, Jan 2026 | sdLM framework, 336 doctrine publications, 127 historical counterfactuals

Uncertainty-Calibrated Explainable AI for Fetal Ultrasound Plane Classification
arXiv, Jan 2026 | Monte Carlo dropout, deep ensembles, conformal prediction

View all publications on ORCID

Medical AI & Imaging: Explainable AI, Federated Learning, Medical Imaging, Neuroradiology, Fetal Ultrasound, Computer-Aided Diagnosis, Clinical Decision Support

Computational Biology: Proteomics, Multi-Omics Integration, Systems Biology, Genomics, Bioinformatics, Mass Spectrometry, Gene Regulatory Networks

Machine Learning Theory: Reinforcement Learning, Graph Neural Networks, Physics-Informed Neural Networks, Bayesian Methods, Uncertainty Quantification, Continual Learning

Human-AI Collaboration: Interactive Machine Learning, Human-in-the-Loop AI, XAI, Trust and Transparency, Usability Studies

Privacy & Security: Federated Learning, Privacy-Preserving AI, GDPR Compliance, HIPAA Compliance, Differential Privacy

Urban Intelligence: Geospatial Analysis, Climate Resilience, Smart Cities, Urban Planning, Disaster Risk Prediction

Applied AI: Time Series Forecasting, Computer Vision, NLP, Optimization, Vehicle Routing, Energy Efficiency

International Organizations
Association for Computing Machinery (ACM) | IEEE Computer Society | Medical Image Computing and Computer Assisted Intervention Society (MICCAI Society) | International Society for Computational Biology (ISCB)

European Networks
European Public Health Association (EUPHA) | European Society for Medical Oncology (ESMO) | European Association for Bioinformatics in Proteomics (EuBIC) | Nordic Society for Bioinformatics (NSBi)

Research Networks
The International Network for Research on Inequalities in Child Health (INRICH) | School as an Origin of Health Disparity and Adversity (SOHDA) | Society for Longitudinal and Lifecourse Studies (SLLS)

Student Organizations (Past)
AI & Data Science Society (Linköping, Tampere) | IEEE Student Branch Tampere | Robotics Club | MUN | Debate Society

Professional Proficiency: English (C1 - IELTS 8.0/9.0), Finnish (C2 - YKI Level 6, Native), Swedish (C1 - Swedex B2), Danish (B2 - PD3), Norwegian (B2 - Norskprøve 3), Azerbaijani, Turkish, French (B1 - DELF B1)

Limited Proficiency: German

Certifications: IELTS Academic C1 (8.0/9.0) | YKI C2 (Level 6) | Swedex B2 | Prøve i Dansk 3 (B2) | Norskprøve 3 (B2) | DELF B1

• Research Excellence Award 2025 - Technical University of Denmark
• University Merit Scholarship 2025 - University of Luxembourg
• Excellence Scholarship 2024 - Linköping University
• President's Medal 2024 - Tampere University
• Faculty Excellence in Teaching Award Nomination 2025 - Linköping University
• Dean's List - DTU (2025), Luxembourg (2025), Linköping (2024-2026), Tampere (2021-2024)
• 50,000+ medical scans processed - Skolyn clinical pilots across Nordics and DACH
• $2M Seed funding - Skolyn (positioning for Series A)
• 2 publications - Nature Medicine, Patterns (FCAI)
• 3 conference presentations - MICCAI 2025, ECR 2026, Nordic AI in Medicine Summit
• LinkedIn Career Boost - Google Health AI, Uppsala University positions

Skolyn AI Platform
127 pathological indicators per scan in <3 seconds | 95%+ diagnostic accuracy | Full XAI architecture | CE Mark, ISO 13485, FDA 510(k) ready

DTU Proteomics Pipelines
10+ TB data processed annually | 30% computation runtime reduction | 6 HPC environments (SLURM) | Nextflow/nf-core standards

Uppsala Medical Imaging
3+ image processing pipelines | 25-30% manual annotation reduction | 95%+ AUC diagnostics | Federated learning across 5 Swedish hospitals | 0.9+ accuracy under GDPR

Google Health AI Evaluations
50+ board-certified clinicians | 1M+ simulated patient interactions | 18% bias reduction | 0.82 inter-rater reliability

FCAI Federated Learning
AUC improvement from 0.91 to 0.95 | 5-hospital federation in Finland | GDPR/HIPAA compliant | 40% training data increase without centralization

ERC Policy Analysis
12,000+ ERC projects analyzed | 75,000 researchers | 200,000+ publications | 2,200 patents | 27 EU Member States

Linköping Teaching
6 courses delivered | 150+ students per semester | 25+ Jupyter notebooks | 90%+ student satisfaction | 10+ supervised projects

Primary Email: oyli@dtu.dk
Affiliation: Department of Applied Mathematics and Computer Science (DTU Compute)
Institution: Technical University of Denmark
Address: Richard Petersens Plads, Building 324, 2800 Kongens Lyngby, Denmark

Additional Affiliations:
University of Luxembourg | Uppsala University | Skolyn | Linköping University

My work focuses on building AI systems that are accurate, transparent, and aligned with human values. I believe the future of AI in medicine and science depends on systems that clinicians and researchers can understand, audit, and trust.

Core Research Philosophy:

• Explainability First: Developing interpretable models that reveal reasoning processes
• Privacy by Design: Building federated and privacy-preserving systems from the ground up
• Human-Centered AI: Creating systems that enhance rather than replace human expertise
• Clinical Validation: Ensuring AI systems meet real-world safety and efficacy standards
• Uncertainty Awareness: Quantifying and communicating model confidence reliably
• Reproducible Research: Maintaining rigorous documentation and open science practices

I am particularly interested in advancing:

• Explainable AI for high-stakes medical decisions
• Federated learning frameworks for multi-institutional research
• Uncertainty-aware models with calibrated confidence
• Human-in-the-loop systems for collaborative decision-making
• Multi-omics integration for systems biology
• Trustworthy AI deployment in clinical and urban environments

Dual PhD Research:

• Human-XAI collaboration for improved fetal ultrasound imaging (DTU)
• Computational models of cellular signaling for predictive biomedicine (Luxembourg)

Industry Innovation:

• Explainable clinical AI at scale (Skolyn)
• Urban intelligence and climate resilience (Azerbaijan projects)

Open Research Questions:

• How can we design XAI systems that clinicians actually trust and use?
• What are the fundamental trade-offs between model accuracy and interpretability?
• How can federated learning scale to real-world clinical consortia?
• What frameworks ensure AI systems remain calibrated under distribution shift?
• How do we build multi-omics models that generalize across populations?

I welcome research collaborations, consulting engagements, and speaking opportunities on:

Research Areas:

• Explainable AI and uncertainty quantification
• Federated learning and privacy-preserving ML
• Medical imaging and computer-aided diagnosis
• Multi-omics data integration and systems biology
• Human-AI interaction and trust in AI systems
• Urban AI and climate resilience

Industry Applications:

• Clinical AI deployment and regulatory strategy
• Proteomics and genomics pipeline development
• Healthcare data infrastructure and governance
• AI safety and evaluation frameworks

Academic Engagement:

• Guest lectures and seminars
• PhD/MSc co-supervision
• Grant proposal collaboration
• Peer review and editorial work

For inquiries: oyli@dtu.dk