Experience
- Research Scientist (2023-Present)
- Vatic Labs
I decided to join Vatic Labs to focus on the frontiers Digital AGI with sandbox of quantitative finance challenges.
- Research Scientist (2021-2023)
- Azure AI Research
I was involved with AI Alignment with RLHF / RLFAI (Reinforcement Learning from Human or AI Feedback). In 2021, I also supported the Azure Personalizer Service (contextual bandits for Personalization / Recommendations).
- Research Engineer / Researcher (2019-2023)
- Microsoft Research
I was part of a wonderful team at MSR, focusing on the Embodied Intelligence pillar, trying to design system with reliable Offline pretraining (through Offline RL, Meta-RL and other interactive representation learning methods) and downstream finetuning, involving Skill learning and discovering for fast online adaptation.
- Research Engineer (2019-2021)
- Microsoft Research
I was a Research Engineer in the Cloud Efficiency group, where contributed to several “Self-driving datacenter” scenarios on Resource Central (ML prediction-serving System for Azure).
- Software Engineer (2017-2019)
- Azure Data AI
Worked on Automatic Configuration Tuning for PostgreSQL DBMS using Deep Reinforcement Learning and SKU recommendation system (best tier/size configuration) of Azure SQL Database to accommodate on-premises workloads
Education
- MSc in Image, Vision and Machine Learning (2014-2017)
- École Polytechnique (France)
Dissertation: Using Machine Learning to build P300-based Brain Computer Interfaces for Gaming
- BSc Computer Software Engineering (2011-2016)
- Instituto Tecnológico de Aeronáutica (Brazil)
Dissertation: A Framework for games controlled by the brain through P300 Visual Event-Related Potentials
- Medical School (2008-2011)
- University of Brasília (Brazil)
I also studied three years of MedSchool, when I spent some time studying the Brain, Human Intelligence and Cognitive Sciences, from where I try to get inspiration and inductive biases for modeling AI agents.