Summary
Arthur Scardua is a Lead Machine Learning Engineer with a PhD in Physics and nine years of experience building production-grade ML systems and platforms in fintech. At Nubank he architects the company-wide ML lifecycle—development, training, deployment, inference and debugging—and led the creation of an Artifact Store that added governance, versioning and security for models used by millions of customers. His background in Bayesian inference, statistical modeling and high-performance scientific computing (MPI/OpenMP/CUDA) informs pragmatic solutions for scalable recommendation systems and reproducible experimentation with MLFlow. Comfortable across Python, Spark, Scala, Kubernetes and Kubeflow, he blends deep research instincts with platform engineering to turn complex models into auditable, reliable services. An often-overlooked strength is his ability to translate rigorous numerical methods from physics into robust data-product infrastructure for large-scale marketing and financial use cases.
9 years of coding experience
5 years of employment as a software developer
Bacharelado, Física, Bacharelado, Física at Universidade Federal do Espírito Santo
Mestrado, Cosmologia, Mestrado, Cosmologia at Centro Brasileiro de Pesquisas Físicas