<aside>
<img src="/icons/link_gray.svg" alt="/icons/link_gray.svg" width="40px" /> š adrian.wilkinscaruana.com
āļø [email protected]
š +61 407 381 557
</aside>
<aside>
<img src="/icons/info-alternate_gray.svg" alt="/icons/info-alternate_gray.svg" width="40px" /> Adrian is an experienced machine learning researcher and practitioner. From MLOps to R&D, he brings industry and academic expertise to a broad range of data-driven projects.
</aside>
Education
Ph.D. in Computer Science
University of Technology Sydney
Jan 2021 -Ā Mar 2024Ā
Thesis:Ā Modelling population-level treatment patterns in patient-level administrative health records
B.Eng. in Mechatronic Engineering
University of Technology Sydney
****Mar 2014 - Jul 2019
Thesis:Ā Hardware-implemented deep learning algorithms for quad-tree partitioning in HEVC
Talks
2022: Department of Health, Australian Government. Title: Abstract Pattern Recognition in Administrative Health Data
2022: Australasian Joint Conference on Artificial Intelligence (paper)
2021: Childrenās Cancer Research Unit, Westmead Hospital. Title: Discovering Treatment Patterns in EHRs
2018: Australasian Conference on Robotics & Automation (Paper)
References
A comprehensive list of professional and academic references is available upon request.
Experience
Nearmap LTDĀ | Machine Learning EngineerĀ
Jan 2019 - Aug 2022, and Jul 2023 - Present
As a full-stack ML Engineer, I have contributed to all aspects of Nearmapās AI capability, from model R&D to production system design and implementation. Specific contributions include:
- Design, training, and evaluation of deep neural networks
- Design of label sampling projects and A-B tests.
- Extensive production system engineering.
- Mentoring of junior data scientists and engineers.
Atlassian Corporation PLCĀ | Machine Learning ScientistĀ
Aug 2022 -Ā Jul 2023
- I build machine learning models that analyse user and customer behavioural data to derive business and marketing insights.
- I co-develop, demonstrate, and present Atlassianās internal ML tooling to help drive ML adoption and fluency within the company.
FreelanceĀ | Researcher, Lecturer, Writer
- Unbox ResearchĀ [2022 - Present]: I translate emerging ML research into understandable blog posts.Ā You can read them here.
- University of Technology Sydney [2021 - 2022]: Lecturing, tutoring, and admin for 32130 Fundamentals of Data Analytics.
- Big Sand [2021] is a sci-fi virtual band exploring new ways to interact with fans. I investigated how AI & LLMs fit into this goal.
Blackmagic Design PTY LTDĀ | Research EngineerĀ
Dec 2017 ā Dec 2018
- I undertook ML R&D for real-time FPGA-based video encoding.
- Replaced rule-based, heuristic algorithms with faster, more precise neural network-based alternatives.
- Wrote and reviewed code for design and verification, and developed automated system integration processes.
Projects
Skills
- AI / ML, Data Science, MLOps, software engineering, data visualisation
- Programming languages: Python, Bash, C/C++, SQL, JavaScript, VHDL, Haskell
- Packages: PyTorch, Scikit-Learn, TensorFlow, Pandas, Numpy, PySpark
- Tools: Git, Docker, Linux/Unix, Tmux, Databricks, AWS, Kubernetes, Kubeflow
Publications
Links: My arXiv papers, my ORCiD, and my Google Scholar profile.
- A. Wilkins-Caruana, M. Bandara, K. Musial, D. Catchpoole, and P. J. Kennedy. Inferring actual treatment pathways from health records. Journal of Biomedical Informatics, 2023. https://doi.org/10.1016/j.jbi.2023.104554
- A. Caruana, M. Bandara, K. Musial, D. Catchpoole, and P. J. Kennedy. Machine learning for administrative health records: A systematic review of techniques and applications. Artificial Intelligence in Medicine, 2023. https://doi.org/10.1016/j.artmed.2023.102642
- A. Caruana, M. Bandara, D. Catchpoole, and P. J. Kennedy. Beyond topics: Discovering latent healthcare objectives from event sequences. InĀ AI 2021: Advances in Artificial Intelligence, pages 368ā380. Springer International Publishing, 2022. https://doi.org/10.1007/978-3-030-97546-3_30
- A. Caruana and T. Vidal-Calleja. Very low complexity convolutional neural network for generating quadtree structures.Ā Australasian Conference on Robotics and Automation, 2018.