Charles Poli

Staff Data Scientist at Chainalysis

Theoretical physicist turned data scientist, and author of eight open-source Python toolkits that teach science, engineering and finance through the history of their breakthroughs.

Open source

Scientific toolkits

PayPal open source

Notes

All notes · RSS

About

I'm a Staff Data Scientist at Chainalysis, where I apply machine learning to blockchain analytics. Before that I built AI-driven trading systems at Cambridge Quantum Computing and fraud-detection tooling at PayPal, where I open-sourced Iguanas and Gators.

I started out as a theoretical physicist. My PhD and postdoctoral research in France, Italy, Mexico and the UK covered wave chaos, random matrix theory and topological states, in close collaboration with experimental groups. I still approach data problems the same way: build a simple model, test it against reality, and make sure the answer holds up.

Selected papers

  1. Selective enhancement of topologically induced interface states in a dielectric resonator chainC. Poli, M. Bellec, U. Kuhl, F. Mortessagne, H. SchomerusNature Communications 6, 6710 (2015)
  2. Nearest level spacing statistics in open chaotic systems: Generalization of the Wigner surmiseC. Poli, G. A. Luna-Acosta, H.-J. StöckmannPhysical Review Letters 108, 174101 (2012)
  3. Statistics of resonance states in open chaotic systems: A perturbative approachC. Poli, D. V. Savin, O. Legrand, F. MortessagnePhysical Review E 80, 046203 (2009)

Full list on Google Scholar.

PhD thesis

Wave chaos in open systems: a non-Hermitian random matrix theory approachUniversité de Nice-Sophia Antipolis, 2009. Advisors: F. Mortessagne and O. Legrand. Written in French.

Experience