Time:  Wednesday, 7th July, 2022,  09.00-10.30  
Speaker: Sean Murphy
 

Abstract: We describe the process of maintaining and improving large-scale neural recommender in an organisation as big and complex as Amazon and how that differs from an academic setting. What techniques we found useful and what we didn’t find useful, and what we have yet to explore.

Bio: I have a degree in maths and physics from University College Cork in Ireland. I moved to Edinburgh in Scotland in 2004 where I acquired a master’s in high-performance computing from the University of Edinburgh. I worked in the domain of medical image analysis for 8 years with Toshiba where I acquired my doctorate. I’ve been working as a machine learning scientist in Amazon for the last 8 years. The first 5 of which were in the advertising domain, buying ads in real-time on the internet. For the last 3 years I’ve been working in the recommendations space, building and improving neural recommenders for the Amazon website.