I am a Software Engineer working on AI/ML for drug discovery at GSK. Previously, I was a Software Engineer on the Analytics team at Zume. I received my Masters in Machine Learning at the University College London under the supervision of Tim Rocktäschel and Jakob Foerster as a member of UCL DARK, where I studied deep reinforcement learning. As an undergraduate, I was a member of the Duke Neutrino and Cosmology Group and Duke ATLAS, the Musical Director for Something Borrowed Something Blue, and - in another life - a bassoonist in the Stephen Paulson Bassoon Studio at the San Francisco Conservatory of Music. I became interested in artificial intelligence in 2018, when I received my Bachelors of Science from the Department of Physics at Duke University under the supervision of Kate Scholberg and Alexander Hartemink.

Like my background, my interests are broad. Current high level projects include:

  • productionizing machine learning models for improving early stage drug discovery
  • building cloud native data ingestion tools for engineers
  • reading the Great Books of the Western World

I have done earlier work in:

  • deep reinforcement learning for social dilemmas
  • building data pipelines for large corpuses of data
  • effects of supernova neutrinos on dark matter detectors
  • direct dark matter detection at the Large Hadron Collider (LHC)

Projects

I want to help develop artificial intelligence to solve important problems.  Projects help me hone my skills in software engineering, machine learning, and different tools.

  • PAX: Scalable Opponent Shaping Experiments in JAX. [Paper] [Website] [Code]
  • Velma: Data pipelines for writing abstracts in your academic field. [Code]

Awards

  • NeurIPS Top Reviewer (~7% awarded, 2024)

Contact

Newton Kwan
Email: nk at 2398.com