savvas is a london-based data scientist and consultant.

they operate within the financial and healthcare sectors specialising in natural language processing, statistics, and strategy. previously, they worked with computer vision techniques to understand parkinsonian tremor and simulation methods to model drug efficacy.

they hold degrees in:

• bsc mathematics and statistical science, first class honours, university college london, link
• msc computational statistics and machine learning, distinction, dean's list, university college london, link

as part of their master's dissertation, they researched how natural language models can use structured data (spreadsheets, tables) to answer statements and assess their validity. they developed a mechanism that can process unstructured data (web scraped text + numbers) and enable NLP models to generalise over them.

their research can be found here:

• claim verification using tabular data by savvas pafitis, pasquale minervini, sebastian riedel, link
• network meta-analysis for parametric survival curves by witold wiecek, savvas pafitis, link

savvas can be reached via:

• email
• linkedin
• twitter

savvas loves the arts; from magritte, picasso, and christoforos savva to sun ra, bjork, and hiromi uehara. they listen to tons of music and occassionaly paint and design.