Samson Tan

San Francisco Bay Area United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Samson Tan is a Staff Research Scientist in the Bay Area with nine years of experience advancing NLP and LLM pretraining, robustness, tokenization, multimodality, and evaluation at organizations including Reddit and Amazon. He combines rigorous academic training (PhD, NUS) with production-savvy engineering—owning end-to-end tokenizer pipelines, speeding up decoding by 2700x in a bugfix, and leading cross-org scaling-law experiments that reduced pretraining time. At AWS he helped ship Amazon Titan tokenizers and built metrics and red-teaming tooling later used in Nova production; at Reddit he now leads pretraining science for platform-native LLMs. He is equally comfortable designing statistically sound evaluation frameworks and writing clean, scalable code, and prefers working in collaborative, kind teams. An active open-source contributor, his work spans NLP augmentation, plotting test coverage in matplotlib, and core tensor ops in PySyft, reflecting a rare blend of research publication pedigree and hands-on implementation. Colleagues describe him as curious, creative, and both scientifically and operationally fluent.

9 years of coding experience

1 year of employment as a software developer

Diploma with Merit, Digital Visual Effects, Diploma with Merit, Digital Visual Effects at Nanyang Polytechnic

Semester abroad, Computer Science, Linguistics, Entrepreneurship, Semester abroad, Computer Science, Linguistics, Entrepreneurship at The University of British Columbia

Honours College, Honours College at NUS University Scholars Programme

Doctor of Philosophy, Computer Science, Doctor of Philosophy, Computer Science at National University of Singapore

English, Chinese, Japanese

Github Skills (35)

data-visualizations10

unit-testing10

pytest10

operation10

python10

user-feedback10

testing10

matplotlib10

tensorrt10

machine-learning10

data-visualisation10

plot10

java10

numpy10

javas10

Programming languages (9)

JavaCSSRustJavaScriptGoJupyter NotebookRubyCython

Github contributions (5)

GEM-benchmark/NL-Augmenter

Jun 2021 - Jul 2021

NL-Augmenter 🦎 → 🐍 A Collaborative Repository of Natural Language Transformations

Role in this project:

ML Engineer

Contributions:67 reviews, 15 commits, 1 PR in 1 month

Contributions summary:Samson primarily contributed to the implementation and refactoring of natural language transformations within the `nl-augmenter` repository. They added new perturbation methods focused on English inflectional variations, including question-specific transformations. The user modified existing code to generate multiple perturbed outputs and refactored code for efficiency and maintainability.

nlpinformation-theorysentencedeep-learningmachine-learning

TEAMMATES/teammates

Jan 2017 - Apr 2017

This is the project website for the TEAMMATES feedback management tool for education

Role in this project:

Full-stack Developer

Contributions:12 commits, 42 PRs, 17 pushes in 3 months

Contributions summary:Samson made several contributions to the TEAMMATES feedback management tool, primarily focusing on both front-end and back-end changes. Their work included updating the UI and data display related to feedback entries, specifically modifying the submission count and presentation on the index page. They also refactored code and modified testing configurations related to instructor feedback result pages. Furthermore, the user improved the status messages given to the user and refactored code structure and improved performance.

cssjavascriptteammatesfeedback-systemsproject-website

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.

Request Free Trial