AI & MACHINE LEARNING
How can we make
machine learning more reliable?
That question shapes my research interests, from evaluating LLM and RAG systems to learning from geospatial and environmental data.
Two research directions.
A shared interest in reliability.
Understanding where models work,
where they fail, and why.
01 / RESEARCH INTERESTS
I’m interested in two complementary areas. Each brings its own questions about evaluation, uncertainty, and how models behave outside familiar settings.
I want to understand when an AI system’s answers can be trusted, how well they are supported by evidence, and what happens when that evidence is misleading.
How do we test whether a useful answer is also a well-supported one?
I’m interested in using machine learning to study spatial and environmental data, especially when a model needs to work in places unlike those it learned from.
How well does a model generalise to a different region?
02 / RESEARCH & PROJECTS
My projects span retrieval-augmented generation and geospatial machine learning.
AI / LLMS & RAG
A helpdesk copilot project using retrieval-augmented generation. My interest is in how retrieved information supports an answer and how the reliability of that answer can be evaluated.
GEOSPATIAL / SCIENTIFIC ML
A geospatial soil classification project, connected to my interests in spatial modelling and the use of machine learning with environmental data.
For future research, I’m interested in RAG safety evaluation, remote sensing, and the challenges of applying spatial models across different regions.
03 / ABOUT & EDUCATION
My academic background is in advanced computer science. I’m looking to develop my interests in reliable AI and geospatial machine learning through PhD research.
I’m drawn to questions that need careful testing: whether an answer is supported by its sources, whether a model works on unfamiliar data, and how to communicate uncertainty honestly.
University of Hertfordshire
Through a PhD, I want to build a stronger foundation in machine learning research and contribute to work on reliable intelligent systems or scientific applications of ML.
04 / GET IN TOUCH
I’d welcome a conversation about PhD opportunities, potential supervision, or research in either of these areas.