James' research focuses on Big Data and the use of Machine Learning techniques in Astronomy. His primary goal is to develop fast pipeline tools in order to prepare for first light of the Square Kilometre Array (SKA). Applications include rapid ML assisted source detection in HI datacubes and kinematic parameterisation using physics-aware neural networks. James has experience using data science practices in both the astronomy world and in wider industry, including: PyTorch CUDA accelerated scripting, REST API microservice deployment, containerisation using Docker, AWS cloud computing, and high-performance computing.
Fields of interest
- AI, Machine Learning, and Data Science Practices
- PyTorch CUDA Accelerated Machine Learning
- Radio Astronomy
- Galaxy Kinematics