Research model
MixtureBetaVAE
Mixture-of-encoders β-VAE framework for interpretable representation learning across scientific domains.
View project →Machine Learning Researcher · Scientific Computing · Interpretable AI
Building interpretable AI systems for scientific discovery.
I explore how generative representation learning can reveal meaningful structure in complex data — from retinal disease progression to tropical cyclone genesis.
Research snapshot
Two scientific domains. One representation-learning question.
My research asks how machine learning models can help us reason about complex transitions, instead of simply assigning labels.
Ophthalmic AI
Interpretable modelling of retinal structure using OCT-derived RNFL and GCIPL thickness maps.
Open research area →Climate AI
Spatio-temporal representation learning for tropical cyclone genesis using ERA5 and OWZ data.
Open research area →Core method
MixtureBetaVAE
A mixture-of-encoders β-VAE with shared decoder, designed for interpretable latent spaces, counterfactual traversal, and scientific reasoning.
View methodInteractive preview
Latent spaces should be explored, not just evaluated.
This placeholder demo shows the type of interface we can later connect to real model outputs.
Healthy
Suspect
Glaucoma
Suspect-like transition
The latent point is moving through an intermediate region. Later, this can display real RNFL/GCIPL reconstructions or cyclone tensor frames.
Featured projects
Research code, scientific pipelines, and experimental builds.
HPC pipeline
TC Tensor Builder
OzSTAR-ready workflow for constructing spatio-temporal cyclone seed tensors from OWZ and ERA5 data.
View project →Medical imaging
OCT Processing
Preprocessing and modelling pipeline for RNFL and GCIPL maps used in glaucoma representation learning.
View project →Path
A research journey across code, science, and interpretation.
Nepal
Early curiosity, computing, and problem solving.
Australia
Academic growth and research training.
Machine Learning
Representation learning, generative models, and scientific data.
PhD Research
Interpretable AI for ophthalmic imaging and tropical meteorology.
Writing & notes
Research thinking, technical notes, and build logs.
Let’s connect
Interested in interpretable AI, scientific ML, or research collaboration?
I’m always open to thoughtful conversations around machine learning, scientific computing, research workflows, and ideas that connect AI with real-world data.
Get in touch