I'm a M. Sc. student in Physics and Machine Learning at UFRGS, working mainly with reinforcement learning and evolutionary dynamics, game theory, machine learning applied to physics and physics inspired neural networks. I have had experiences that greatly helped me implement ML solutions into code, such as working in one of the largest Fintech's in LATAM as a backend developer, freelancing and in previous research. Also, I'm currently working as an independent contractor for Outlier AI. On the personal side, I am always eager to learn and do stuff in new and creative ways, trying to push barriers and, hopefully, stumble into something useful along the way.
I develop high speed simulations for large scale games which use mainly reinforcement learning creating unique implementations in C, Python and Bash. I work side by side with highly experienced and respected researchers in the field, researching and finding novel areas of applications in Machine/Deep Learning (M/D L), Game Theory and RL.
I work aiding in AI training for coding in C, Python, Bash and Julia.
I worked as a software developer on Algotools, a service that was used to create trading strategies tailored for each client. Development was quick and demanded high speed and low latency, as our application took information from the stock market and traded algorithmically.
I assisted small companies with mathematics, physics, and programming problems, helped people with the aforementioned subjects through tutoring and communicated directly with clients to understand and meet their needs
I work within the Physics department, working with reinforcement learning in evolutionary dynamics, but also managing Journal Clubs and Book Clubs, studying general Deep Learning and assisting young researchers in my group.
Besides all mandatory classes, I worked within the mathematics departmeny, developing formal mathematical techniques for dealing with probability and statistics. This layed foundations for my future work, mainly with Markov Chains and Processes.
Mangold, Gustavo C. and Mendeli H. Vainstein. "Dynamical Systems and a Quantum Mechanics Interpretation through the lenses of Fuzzy Logic". Lume, UFRGS, 2023.
Mangold, Gustavo C., Heitor Fernandes, and Mendeli H. Vainstein. "Dilution, Diffusion and Symbiosis in Spatial Prisoner's Dilemma with Reinforcement Learning." arXiv preprint arXiv:2507.02211 (2025).
A Study of Dilution and Mobility in the Spatial Prisoner’s Dilemma with Reinforcement Learning. Talk given at Dynamic Days, University of Buenos Aires, 2024.
Effects of Space and Mobility in the Spatial Prisoner’s Dilemma with Reinforcement Learning. Talk given at ENFE, UFPR, 2024.
A Study of Dilution and Mobility in the Spatial Prisoner’s Dilemma with Reinforcement Learning. Poster presentation at StatPhys29, Florence, 2025.