Paulo Henrique de Souza Macedo Arruda

Designed a new algorithm that fully classifies the partition regularity of polynomial equations in two variables. Developed a new algorithm to characterise linear combinations of idempotent ultrafilters, deepening the connection between ultrafilters and Rado’s Theorem for linear equations.

Collaborated with clients across various industries to design, develop, implement, deploy, improve software efficiency, automate, and build visual reports and dashboards. Successful cases include Banco24h (the largest Brazilian ATM provider), Banco Itaú (the largest Brazilian private bank), Bloomberg, Amazon and Ambev. Provided training to client teams to ensure they can maintain, improve, and actively research, learn, and apply new and state-of-the-art technologies to solve business problems

Implemented the statistical framework Item Response Theory, providing accurate evaluations of student proficiency across four Brazilian states. Implemented the school curricula integration process, guaranteeing full integration with the items database and automating the previous manual integration, thus reducing the previous manual loading process from one week to a few seconds.