About me

Hello!

I am a data professional with a decade of experience working on projects across a variety of industries (health data, occupational therapy, transport, data for good, tourism, finance and ESG), in a variety of capacities (from in-house data scientist, to Shiny developer and principal data consultant, to product owner).

My current work could be called LLM-adjacent: while not directly working on training or fine-tuning at the minute, I am involved in the applications/use cases side of how these models can catalyse organisational change under the right conditions.

I enjoy public speaking and have also been heavily involved with community events in the past, having organised DataTech19 and the Edinburgh R meetup (EdinbR) for a number of years. Check out my Engagements page for a lineup of contributed talks.

Background

Prior, my academic background is in psychology, emotion and research methods, and includes a PhD in Psychology awarded by The University of Edinburgh with no corrections. My doctoral research investigated if various emotion-generating stimuli used in lab settings (including VR & at the time, the recently launched Oculus Rift DK1) could approximate emotional states occurring in daily life, as measured using a bespoke Android phone app. As part of my doctoral research I also tested if model-based cluster analysis can serve as a computational model for how humans create (fuzzy) emotional categories.

Publications

• Constantinescu, A. C., Wolters, M., Moore, A., & MacPherson, S. E. (2017). A cluster-based approach to selecting representative stimuli from the International Affective Picture System (IAPS) database. Behavior Research Methods, 49(3), 896-912.