Experience
ML/AI Research Intern
Bureau of Meteorology (Dec 2024 - Present)
- Investigating the feasibility of using empirical normal modes as a significantly more parameter efficient and interpretable means of forecasting with ML/AI.
- Developed mathematical theory and implemented scalable solutions for evaluation on the NCI Gadi supercomputer.
- Supervised by Gilbert Brunet (former Chief Scientist at BoM), Catherine de Burgh-Day and Craig Bishop.
- Aiming to turn results into a paper over 2025.
Member of ML/AI Research Group
Monash University (Mar 2024 - Present)
- Member of A/Prof. Mehrtash Harandi’s group. Informal member while I complete my undergraduate studies.
- Interests include embedding prior knowledge into models (Bayesian ML, PINNs, reasoning LLMs, etc.), neurosymbolic AI and interpretability (SLT, GDL, etc.).
- Piqued the interest of researchers at Princeton and OpenAI. The latter offered to discuss my ideas in person.
Google Summer of Code Contributor
NumFOCUS, PyMC (May 2025 - Sep 2025)
- Implementing Integrated Nested Laplace Approximations (INLA) into the PyMC library (Statistical ML). More info here and here.
- Developing efficient, maintainable Python code for an open-source repository.
- Working with probabilistic programming (PyMC) and an autodifferentiable backend (PyTensor).
- Supervised by Rob Zinkov (University of Oxford), Theo Rashid (Amazon) and Colt Allen (PyMC Labs).
Research Intern
University of Melbourne (Nov 2022 - Mar 2025)
- Conducted research on the behaviour of heat fluxes in tropical cyclones – a poorly understood, data and modelling-heavy field.
- Published in Geophysical Research Letters (Q1 Journal). More info here.
- Was promoted to lead author within just eight months of starting out as a research assistant.
- Experience with working with gigabytes of noisy data.
- Made data analysis up to 48 times faster by introducing Python multiprocessing.
Simulation and Machine Learning Engineer
Monash High Powered Rocketry (Apr 2022 - Feb 2025)
- Led the research and development of ANDROMEDA, a data-driven tool which sped up fluid simulations from taking hours to just minutes.
- Won second place for research and modelling internationally (Spaceport America Cup, Charles Hoult Award).
- Vice lead of the Dynamics section during the 2023-2024 management cycle.
- Oversaw and managed section R&D projects.
- Reported to CTO at technical meetings.
- Developed and ran pipelines on the MonARCH cluster (slurm).
- 2nd largest contributor to the GitLab by Git contributions as of writing.
Education
BEng (Hons) Aerospace Engineering, Minoring in Artificial Intelligence
Monash University (Feb 2022 - Nov 2026)
- WAM 81.5 (High Distinction)
- Selected cohort rankings:
- 3rd of ~200 in ENG2005 Advanced Engineering Mathematics.
- 8th of 197 in ECE4179 Neural Networks and Deep Learning
- 2nd of 57 in MAE3401 Aerodynamics 2
- 5th of 42 in MAE3404 Flight Vehicle Dynamics
- Top 5% in MEC3456 Engineering Computational Analysis
University Extension Program (Physics)
University of Melbourne, (Feb-Nov 2021)
- Selected as one of around 20 students statewide based on academic performance to undertake two first year physics units during my final year of high school.
- H1 Honours (High Distinction).
Publications
Novomestsky, M., Voermans, J. J., & Babanin, A. V. (2025). In situ measurements of sensible heat fluxes in a tropical cyclone. Geophysical Research Letters, 52, e2025GL115842. https://doi.org/10.1029/2025GL115842
Teaching
Competitions
International Young Physicists’ Tournament (IYPT), 2021
- Captain of the Australian team.
- Represented Australia on the world stage by leading the national team in an international physics Olympiad.
- Researched and presented on university-level theory including Lagrangian mechanics whilst still in high school.
- Managed and advised teammates amidst researching difficult theory and working with a tight schedule.
Two-time ASO Physics Olympiad Distinction, 2020 and 2021
