Luis Pineda
Research Engineer at Facebook AI
Montreal, Quebec, Canada
Summary
🤩
Rockstar🎓
Top SchoolLuis Pineda is a research engineer with 12 years of experience building algorithms for autonomous systems that make safe, resource-aware decisions under uncertainty. With a PhD from UMass Amherst and roles at Facebook AI, Microsoft, and Google, he blends theoretical work in planning and optimization with hands-on software development in C++ and ML. He has contributed to high-profile open-source projects like FacebookResearch's MBRL-lib and Theseus, improving model save/load robustness, SO3 support, and robust loss functions for differentiable optimization. His background spans applied engineering in oil production optimization, graduate-level teaching, and scalable continual-planning research—bringing rare depth across academic, industrial, and applied domains. Colleagues know him for translating rigorous mathematical approaches into reliable, production-ready systems.
12 years of coding experience
14 years of employment as a software developer
M.Sc. Applied Computing, M.Sc. Applied Computing at Universidad del ZuliaDoctor of Philosophy (PhD) Computer Science, Doctor of Philosophy (PhD) Computer Science at University of Massachusetts Amherst
English, Spanish
Github Skills (10)
pytorch10
machine-learning10
differentiable-programming10
nonlinear-optimization10
python10
reinforcement-learning10
linear-algebra9
automatic-differentiation9
unit-testing8
software-design8
Programming languages (3)
Github contributions (5)
facebookresearch/theseus
Dec 2021 - Jan 2023
A library for differentiable nonlinear optimization
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:771 reviews, 175 commits, 447 PRs in 1 year 1 month
Contributions summary:Luis made several contributions to the differentiable nonlinear optimization library. Their commits included implementing support for SO3, refactoring for improved code structure and clarity, and adding a class for an easier construction of the Theseus layer. They also implemented the Huber loss and its gradients for use in robust cost functions. The user appears to be focused on expanding and improving the core functionality of the library.
pytorchroboticsdifferentiablebilevel-optimizationimplicit-differentiation
facebookresearch/mbrl-lib
Aug 2020 - Jan 2023
Library for Model Based RL
Role in this project:ML Engineer
ML EngineerContributions:7 releases, 77 reviews, 527 commits in 2 years 5 months
Contributions summary:Luis primarily contributed to the model-based reinforcement learning library by adding notes clarifying usage of existing models. Their work also included fixing bugs in tests, such as resolving an issue in the test algorithms where overrides were not setting up correctly. The user made changes to the core files of models within the library, specifically fixing bugs within the save/load methods of existing models, further enhancing the functionality of the existing models.
pytorchmodel-basedmodel-based-reinforcement-learningreinforcement-learningmujoco
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.