Aloysio Galvão Lopes
Engineer at Enter · PhD in Computer Science
AI engineer at Enter — Latin America's first AI unicorn. PhD from Ecole Polytechnique, where I researched real-time probabilistic safety for autonomous vehicles.
Experience
Enter — São Paulo, Brazil
- Building AI products at Latin America's first AI unicorn.
Polytechnique's Computer Science Lab (LIX) — Palaiseau, France
- Researched real-time probabilistic safety for AI methods applied to autonomous vehicles.
- Developed a statistical method for ML-based safe AV trajectory forecasting (published in PMLR).
- Designed a framework for real-time trajectory planning with provable theoretical safety guarantees.
Kili Technology — Paris, France
- Deployed state-of-the-art image segmentation models in a Kubernetes GPU cluster.
- Improved superpixels image segmentation speed by 4× leveraging C++, CPU cache, and CUDA.
Amazon Web Services (AWS) — Berlin, Germany
- Developed a DynamoDB database failure simulation tool and recovery system in Ruby.
Research
Probabilistic Safety for ML-Based Autonomous Vehicle Trajectory Forecasting
Proceedings of Machine Learning Research (PMLR) · 2024
A statistical method for safe autonomous vehicle trajectory forecasting with provable theoretical safety guarantees.
Read paper ↗Real-Time Probabilistic Safety for AI Methods Applied to Autonomous Vehicles
Ecole Polytechnique · 2025
Full dissertation on probabilistic safety frameworks for real-time AI decision-making in autonomous systems.
Read thesis ↗Education
Ecole Polytechnique — Palaiseau, France
Institut Polytechnique de Paris
Graduated with highest honors · GPA 3.8
Skills