Professional on Paulo H. Barchi

Industry & AI

Mon, 01 Jan 0001 00:00:00 +0000

Selected Experience

Staff AI/Data Engineering @bol
Leading technical strategy, architecture, and design for AI and data systems.

Machine Learning Engineering @bol: Recommendations, Search & Logistics
Developed scalable machine learning pipelines, focusing on recommendations, search relevance, and logistics optimization.

AI for Gaming @Wildlife Studios: Reinforcement Learning & Recommendations
Designed and deployed reinforcement learning models for in-game recommendations and player engagement systems.

Super-Resolution & Camera Applications @Instituto Eldorado
Engineered low-latency image processing and neural super-resolution algorithms for smartphone camera hardware.

PhD: AI applied to Astrophysics

Mon, 01 Jan 0001 00:00:00 +0000

Applied machine and deep learning techniques to automate the morphological classification of galaxies at scale (2016-2020)
📍 National Institute for Space Research (INPE), Brazil

Thesis

Machine and deep learning applied to galaxy morphology
🔗 INPE’s digital library

Main Publications

Machine and Deep Learning Applied to Galaxy Morphology — A Comparative Study
P. H. Barchi, R. R. de Carvalho, R. R. Rosa, R. Sautter, M. Soares-Santos, B. A. D. Marques, E. Clua, T. S. Gonçalves, C. de Sá-Freitas, T. C. Moura
Astronomy and Computing, Vol. 30, 100334 (2020)
🔗 ScienceDirect / Journal Article | arXiv:1901.07047

Master: AI & Never-Ending Learning

Mon, 01 Jan 0001 00:00:00 +0000

Explored Machine reading and ontology extension to enable systems to continuously read unstructured web text and use that information to expand and refine semantic ontologies (never-ending learning).
📍 Federal University of São Carlos (UFSCar), Brazil

Dissertation

Expansão de Ontologia através da Leitura de Máquina sem fim
Ontology Extension through Never-Ending Machine Reading
🔗 UFSCar’s digital library

Main Publications

Never-ending ontology extension through machine reading
P. H. Barchi, E. R. Hruschka
14th International Conference on Hybrid Intelligent Systems, pp. 266-272 (2014)
🔗 IEEE Xplore