CV

Curriculum vitae of Andy Tai, Postdoctoral Researcher at the Institute for Artificial Intelligence in Medicine (IKIM), University Hospital Essen.

Contact Information

Name Man Yeung (Andy) Tai
Professional Title Postdoctoral Researcher, Institute for Artificial Intelligence in Medicine (IKIM), University Hospital Essen
Email man.tai@uk-essen.de
Location , Köln, North Rhine-Westphalia
Website https://andytai7.github.io/Andy-Tai

Professional Summary

My research develops federated and privacy-preserving machine learning for trustworthy AI in medicine: methods that let clinical models learn from distributed hospital data without centralizing patient records, with explicit accounting of communication cost, privacy, and reliability. It builds on a PhD in neuroscience in which I applied machine learning to addiction psychiatry, including overdose risk prediction and clinical decision support.

Experience

  • 2026 - Present

    Essen, Germany

    Postdoctoral Researcher
    Institute for Artificial Intelligence in Medicine (IKIM), University Hospital Essen
    Jointly supervised by Prof. Jens Kleesiek (IKIM) and Prof. Michael Kamp (Kamp Lab, Lamarr Institute, TU Dortmund). Federated learning for medical data in the FLIP-IT project (NEXT.IN.NRW-funded), with Dr. Moon Kim.
    • FLIP-IT: federated chronic kidney disease risk prediction (P(CKD stage >= 3)) across general practice clinics using the Flower framework
    • Privacy design: patient-level DP-SGD with SecAgg+ secure aggregation, FedAvg, an RDP privacy accountant with one global (ε, δ) target and per-practice noise calibration, per-round monitoring of global and worst-practice AUROC and sensitivity
    • Governance gate before on-site deployment: membership-inference leakage audit plus clinical and legal review
    • Developing privacy-preserving federated continual learning methods and an open-source benchmark toolkit; work in progress on activation and rank-code communication protocols
  • 2024 - Present

    Vancouver, BC, Canada

    Analyst
    NAI Innovations
    Evaluation of machine learning-driven start-ups in medical cannabis symptom management; algorithmic performance analysis; strategy for AI in neuroscience and psychiatry.
  • 2019 - Present
    Founder / CEO
    Building Blocks (student-led nonprofit)
    Peer mentoring app and community; corporation no. 1176510-8.
  • 2025 - 2026
    Postdoctoral Representative, SEDI Committee
    Department of Statistics, UBC
    Equity, diversity, and inclusion initiatives, policy, seminars, and mentorship programs.
  • 2025 - 2026
    Coordinator, EDI Seminar Series
    Department of Statistics, UBC
    Speaker invitation and scheduling for the department EDI seminar series.
  • 2024 - 2026

    Vancouver, BC, Canada

    Postdoctoral Teaching & Learning Fellow
    University of British Columbia, Master of Data Science, Department of Statistics
    Two-year appointment, completed June 2026. Primary instructor for 10 data science, statistics, and science courses reaching more than 550 graduate and undergraduate students; capstone supervision; curriculum development; educational research.
    • Courses include Data Visualization (DSCI 531), Probability (DSCI 551, STAT 302), Databases and SQL (DSCI 513), and Communication and Argumentation (DSCI 542)
    • PrairieLearn assessment and rubric design; Quarto-based course websites; lecture materials on data visualization (Altair, ggplot2), simulation, and AI literacy
    • Evaluator for MDS capstone presentations with industry partners including Clarius Mobile Health and VRIFY Technology
    • Department of Statistics EDI Seminar Series coordinator; SEDI Committee postdoctoral representative (Sep 2025 - Jun 2026)
  • 2025 - 2025
    Judge, Science Case Competition 2025
    UBC Faculty of Science (SCI TEAM)
    Evaluated team presentations on BC wildfire resilience strategies.
  • 2025 - 2025

    Vancouver, BC, Canada

    Data Science Consultant (freelance, part-time)
    Clause Technology
    Built HAVA, an agentic system for detecting illegal fishing and maritime criminal behaviour, with Nautical Crime Investigation Services (NCIS): automated web scraping, NLP incident classification, and LLM-assisted de-duplication (BERT/NER extraction of vessels, locations, and organizations; fuzzy matching) in a Dockerized, cloud-ready MongoDB/PostgreSQL architecture. Developed a computer vision and OCR pipeline for complex PDF tables (rotations, multi-table pages), including image preprocessing and variable extraction.
  • 2025 - 2025
    Mental Health Ministry Coordinator
    Tenth Church
    Monthly seminars; Sanctuary Course (8 weeks).
  • 2025 - 2025
    Member, Selection Committee
    MasterCard Foundation and UBC
    Reviewed approximately 150 scholarship applications for a fair and inclusive selection process.
  • 2024 - 2024
    Panelist, National Undergraduate Research Conference
    UBC Neuroscience Club
    323 attendees, 41 research presentations.
  • 2024 - 2024

    Vancouver, BC, Canada

    Data Scientist
    Concussion RX
    Machine learning and statistical modelling for concussion subtype analysis; clinical onboarding documentation for the ConcussionRX and Llama 3 toolchain.
  • 2021 - 2024
    Associate Editor
    URNCST Journal
    Mentored undergraduate research teams across natural and clinical sciences.
  • 2019 - 2024

    Vancouver, BC, Canada

    Teaching Assistant
    University of British Columbia (MDS, Computer Science, Business Analytics, Medicine)
    MDS courses: DSCI 522, 523, 524, 531, 532, 541, 551, 552, 553, 571, 573; BAIT 507, 509, 580A; CPSC 121, 322; DSCI 320; NRSC 501. MDS TA Award (2021/22).
  • 2019 - 2023
    CS:GO Director
    UBC Esports
  • 2019 - 2021
    Member
    Walter Gage Toastmasters
    Communication and leadership.
  • 2019 - 2020
    VP Student Life and First-Year Representative
    Neuroscience Graduate Student Association, UBC
  • 2019 - 2020
    Volunteer
    Coast Mental Health, Vancouver
    Mentorship and activities.
  • 2018 - 2019

    Vancouver, BC, Canada

    Research Assistant
    Addiction and Concurrent Disorders Group, Institute of Mental Health, UBC
    E-mental health: WalkAlong; RAMP (Risk Assessment and Management Platform); conference organizing.
  • 2018 - 2018

    Vancouver, BC, Canada

    Research Assistant
    BC Children's Hospital / CFRI, UBC
    Microglia morphology in bipolar disorder and schizophrenia cortical tissue (Beasley Lab).
  • 2018 - 2018

    Vancouver, BC, Canada

    Research Analyst
    NAI Interactive Ltd.
    Biotech pipeline data mining and research.
  • 2017 - 2018

    Toronto, ON, Canada

    Lab Technician
    SickKids, PGCRL (Josselyn Lab)
    Epitranscriptomics; optogenetics; operant behavior; DNA construct work; animal perfusions and imaging.
  • 2016 - 2017

    Toronto, ON, Canada

    Research Assistant
    Toronto Western Hospital, MDPU (McIntyre)
    Big data and ML in psychiatry; meta-analysis on THC and cognition; first-author ML review.

Education

  • 2019 - 2024
    PhD
    University of British Columbia
    Neuroscience (graduate program with fast-track transfer from MSc)
  • 2012 - 2017
    Honours BSc
    University of Toronto
    Neuroscience (major); Environmental Science and Religion (minors)

Awards

  • 2022
    Mitacs Globalink Research Award
    Mitacs Canada (2022-2023, $6,000 stipend and research costs)

    Visiting research at University of Sydney, Brain & Mind Centre (Oct 2022 - Jan 2023): machine learning synergy for digital mental health with Prof. Ian Hickie and Dr. Frank Iorfino.

  • 2022
    MDS Teaching Assistant Award (2021/22)
    UBC Master of Data Science

    Outstanding TA contributions across multiple MDS courses.

  • 2021
    President's Academic Excellence Initiative PhD Award
    University of British Columbia (2021-2023)

    UBC program recognition for PhD research contributions.

  • 2021
    UBC Research Day People's Choice (Lightning Talk and ePoster)
    University of British Columbia

    AI/ML overdose risk modeling.

  • 2019
    Faculty of Medicine Graduate Award
    University of British Columbia

  • 2019
    NIDA Travel Award
    NIDA ($500 USD, 2020 NIDA International Forum)

Publications

Skills

Programming: Python, R, SQL, JavaScript, React (basic)
Machine Learning: PyTorch, TensorFlow, scikit-learn, BERT/transformers, NLP
Federated Learning: Flower framework, federated pipeline design, communication-efficient distributed training
LLM Tooling: LiteLLM gateway configuration for agentic coding tools, LLM-assisted data pipelines
Data and DevOps: PostgreSQL, MongoDB, Docker, AWS (basics), Git
Statistics: Cumulative link mixed models, Stuart-Maxwell tests, meta-analysis (Covidence), SPSS, SAS (basics)
Teaching Tools: PrairieLearn, Quarto, JupyterLab, RStudio, Altair, ggplot2

Languages

English : Native
Cantonese : Fluent
Mandarin : Conversational
German : Beginner

Interests

Methods: Federated and continual learning, Privacy-preserving and communication-efficient distributed training, Deep learning, Clinical prediction and risk modelling, Natural language processing and large language model tooling, Statistical modelling and meta-analysis
Applications: Federated learning for medical data, Addiction psychiatry and the overdose crisis, Digital and youth mental health, Clinical decision support, Maritime surveillance, Data science education and AI literacy

Projects

  • 2026 - Present
    FLIP-IT, Federated learning for medical data

    IKIM, University Hospital Essen; NEXT.IN.NRW-funded. Federated Learning in Practice Networks, co-funded by the European Union and North Rhine-Westphalia, builds infrastructure for decentralized training of medical AI across general practices in NRW (consortium: IKIM, docport, Flower Labs, KVWL, KVNO; project lead: Dr. Nicolas Conze). I build federated chronic kidney disease risk prediction (P(CKD stage >= 3)) on the Flower framework, with a privacy design of patient-level DP-SGD and SecAgg+ secure aggregation, FedAvg, an RDP privacy accountant with one global (ε, δ) target and per-practice noise calibration, per-round monitoring of global and worst-practice AUROC and sensitivity, and a governance gate with membership-inference leakage audit and clinical and legal review before on-site deployment. With Prof. Jens Kleesiek, Prof. Michael Kamp, and Dr. Moon Kim. Project site: https://www.flipit-research.network/

  • 2026 - Present
    Brain-inspired federated continual learning for healthcare AI

    Privacy-preserving federated continual learning algorithms (federated synaptic consolidation, federated hippocampal replay, predictive-coding federated learning) that let clinical models learn continuously from distributed hospital data without catastrophic forgetting; open-source benchmark toolkit.

  • 2026 - Present
    Activation and rank-code communication protocols for federated learning

    Federated protocols in which clients communicate activation or rank codes (which units fire, and in what order) instead of model parameters, with pre-registered control arms and explicit communication-bitrate accounting.

  • 2019 - 2024
    RAMP, Risk Assessment and Management Platform

    Health Canada, Substance Use and Addictions Program. $1,407,790, Co-Investigator. An online overdose risk score with tailored feedback.

  • 2025 - 2025
    HAVA, Maritime criminal detection system

    With Clause Technology and NCIS. Agentic system for detecting illegal fishing and maritime criminal behaviour: automated web scraping, NLP incident classification, BERT/NER entity resolution of vessels, locations, and organizations, LLM-assisted de-duplication, and a Dockerized MongoDB/PostgreSQL deployment.

References

  • Prof. Jens Kleesiek

    Postdoctoral co-supervisor (joint); Institute for Artificial Intelligence in Medicine (IKIM), University Hospital Essen

  • Prof. Michael Kamp

    Postdoctoral co-supervisor (joint); Lamarr Institute for Machine Learning and Artificial Intelligence, TU Dortmund University

  • Dr. Reinhard Michael Krausz

    UBC-Providence Leadership Chair for Addiction Research; Professor of Psychiatry

  • Dr. Raymond Ng

    Professor, Department of Computer Science, University of British Columbia

  • Dr. Varada Kolhatkar

    Associate Professor of Teaching in Computer Science; Co-Director, Master of Data Science (Vancouver)