Pablo Castro
Staff Research Scientist at Mila - Quebec Artificial Intelligence Institute
Ottawa, Ontario, Canada
Summary
🤩
Rockstar🎓
Top SchoolPablo Castro is a Staff Research Scientist at Google DeepMind and long-time Google researcher specializing in reinforcement learning, probabilistic planning, and machine learning for music and creativity. With a PhD from McGill and over a decade of applied research and engineering experience, he blends theoretical work on bisimulation and MDP/POMDP metrics with production-scale ML systems and tooling. He’s contributed to high-profile open-source projects such as Dopamine and Magenta.js, building research frameworks and interactive browser-based music generation demos. Equally comfortable in low-level engineering and academic settings, he teaches as an adjunct professor and is a core industry member at Mila. Based in Ottawa, he’s also a musician, which informs his unique focus on creative ML applications that bridge sound, visualization, and agent behavior analysis.
8 years of coding experience
9 years of employment as a software developer
Colegio Americano de QuitoM.Sc Computer Science, M.Sc Computer Science at McGill University
Github Skills (12)
dqn10
machine-learning10
javascript10
tensorflow10
python10
reinforcement-learning10
matplotlib10
ai9
data-visualisation9
data-visualization9
data-visualizations9
jax6
Programming languages (6)
Github contributions (5)
google/dopamine
Aug 2018 - Sep 2022
Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:2 releases, 140 commits, 5 PRs in 4 years 1 month
Contributions summary:Pablo primarily contributed to the development and improvement of the Dopamine framework, a research tool for reinforcement learning. Their contributions included importing projects, modifying code for specific algorithms like DQN, and adding new features for training, such as in-iteration Tensorboard reporting and providing graph definitions. They also made changes to improve the code's documentation. The user also focused on addressing bugs and improving code maintainability.
reinforcement-learning-algorithmsfast-prototypinggooglereinforcement-learningprototyping
google-research/google-research
Nov 2019 - Dec 2021
Google Research
Role in this project:ML Engineer
ML EngineerContributions:18 commits in 2 years 1 month
Contributions summary:Pablo contributed code related to visualizing and analyzing agent behavior within a reinforcement learning context. They worked on implementing visualizations, including line plots and distributions, to understand agent performance. The user's code integrates with PyGame for rendering and matplotlib for generating plots, which suggests a focus on creating visual tools for the analysis of machine learning models.
googlemachine-learningai
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.