Summary
Marcelo Hartmann is an applied statistician and Academic Research Fellow based in Helsinki with 11 years of experience developing inference algorithms and loss functions for AI, especially leveraging differential geometry and estimating-function theory. He holds a PhD (with distinction) from the University of Helsinki and advanced degrees from top Brazilian universities, and has led research at FCAI and the Academy of Finland on scalable Bayesian methods, Gaussian processes and MCMC on Riemannian embeddings. Marcelo’s work bridges theory and practice—designing score-matching and geometry-informed objectives that aim to reduce storage and computation while improving learning in real-world AI systems. He combines deep expertise in languages from Matlab and C to Julia and Python with applied domains including quantitative ecology, survey engineering and mapping projections. Outside academia he channels creativity through photography, poetry, cooking and even part-time Forró dancing, reflecting a pragmatic yet imaginative approach to problem solving.
11 years of coding experience
7 years of employment as a software developer
Bachelor's degree of Statistics, Time series analysis through bayesian nonparametrics, Second in Class, Bachelor's degree of Statistics, Time series analysis through bayesian nonparametrics, Second in Class at Universidade Estadual Paulista Júlio de Mesquita Filho
Doctor of Philosophy - PhD, Statistics and quantitative ecology, With Distinction, Doctor of Philosophy - PhD, Statistics and quantitative ecology, With Distinction at University of Helsinki
University of São Paulo
German, English, Portuguese