Building clinically useful AI from speech and longitudinal health data.

My doctoral work combines evidence synthesis, diagnostic-accuracy assessment and multilingual deep learning for automatic speech-based depression screening. I also designed and built a working iOS prototype that translates this research into a mobile health application.

Alongside my PhD, I lead an analytical work package within the eLIXIR-BiSL programme, working with more than 30,000 repeated observations from over 18,000 infants across linked NHS datasets. Before research, I worked in data analytics and governance at Deloitte Consulting and Bank of Jilin.

I use Python, PyTorch, R, SQL and Swift, and enjoy collaborating across clinical, academic and data teams to turn complex methods into clear, responsible and reproducible outputs.