Jacob Austin
Member Of Technical Staff at Anthropic
New York, New York, United States
Summary
👤
Senior🎓
Top SchoolJacob Austin is a research-focused software engineer with nine years of experience building ML systems and program synthesis at Google DeepMind and more recently Anthropic. He specializes in scaling large language models and improving ML tooling—contributions include enhancing T5X inference/decoding and implementing game logic for OpenSpiel’s Othello environment. His background blends rigorous CS and mathematics from Columbia with applied research roles (AI Resident, research engineer) and systems work from JPL to NVIDIA. Jacob has a track record of shipping robust code and tests across libraries like python-fire, plus a prize-winning open-source visualization at NASA JPL. He enjoys making practical, well-tested ML and PL tools and brings hands-on expertise in inference pipelines, discrete diffusion models, and back-end engineering. Outside work he balances technical curiosity with endurance running and piano performance.
8 years of coding experience
7 years of employment as a software developer
Bachelor's degree Computer Science Mathematics, Bachelor's degree Computer Science Mathematics at Columbia University
English, French
Github Skills (27)
unit-testing10
c-language10
python10
coregames10
command-line-interface10
machine-learning10
inference10
videogames10
decoding10
deep-learning10
command-line10
nlp10
jax10
cprogramming-language10
cli10
Programming languages (9)
Github contributions (5)
google-research/t5x
Nov 2021 - Oct 2022
Role in this project:
ML Engineer
ML EngineerContributions:12 commits in 11 months
Contributions summary:Jacob primarily contributed to the inference and decoding functionalities of the T5X framework. They modified `infer.py` to support configurable encoder and write functions, enhancing the flexibility of the inference process. Furthermore, the user addressed issues in the decoding process, including fixing topp sampling and adding dynamic parameters, which directly impacts model generation capabilities. These changes indicate a focus on improving the core machine-learning pipeline and user control over model execution.
google-deepmind/open_spiel
Apr 2020 - Apr 2020
OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.
Role in this project:Back-end Developer
Back-end DeveloperContributions:8 commits, 1 PR, 17 comments in 8 days
Contributions summary:Jacob's primary contribution involved implementing the Othello game within the OpenSpiel framework. This included writing the core game logic, defining valid moves, and creating methods for action application and observation. They added new game features and expanded the existing capabilities for Reversi/Othello, and integrated unit tests to validate functionality. The user also focused on refactoring and cleanup throughout the development process, improving code readability and consistency.
cppmultiagentgamespythondatamining
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.