Mathematics Research Seminar Series: Actually Useful AI Systems: From Experiment to Production

Abstract:
It’s an exciting time to work in AI, with new products launched, benchmarks broken, and increasingly complex papers published seemingly every week. The data scientist of today is expected to not only keep up with the literature, but also build actually useful AI systems that solve real business problems.

In this talk, we’ll walk through the process of bringing a large language model from the notebook to production to address a document processing use case. We’ll highlight the importance of solid mathematics and statistical intuition to questions of system design and data modeling. We’ll explore how research looks like in the field of technology consulting, and discover the various hats a data scientist can wear working in industry.

About the Speaker:
Scott Lee Chua is a Machine Learning Consultant at Thinking Machines Data Science solving high-impact problems with AI and data. Previously, he led the modernization of complex machine learning models for a telco, and developed a PDF chatbot for an insurance agency. At present, he builds data extraction pipelines using large language models for a regional financial regulator, and continues to conduct data visualization workshops for professionals from all over the world.

His previous work experience includes SeaMoney Digital Banking in Singapore, Department of Finance in the Philippines, Bank of the Philippine Islands. Scott holds certifications from Microsoft, Amazon Web Services, CFA, among others.

Scott graduated summa cum laude, majoring in Economics and minoring in the Mathematical, Computational and Statistical Sciences from Yale-NUS Singapore in 2020, where he was also commencement speaker and former Student Government president. He finished a master’s degree in Quantitative Economics from the National University of Singapore in 2022. He has co-authored papers in economics and science and technology studies for peer-reviewed journals.

Scott writes fiction and nonfiction, and won the (international) Gabo Prize for Literature in Translation, the Carlos Palanca Memorial Awards for Literature, and the Nick Joaquin Literary Awards. He trains the public in Singapore Math model methods, and designs escape rooms for fun.