Phil Yeres

Software Engineer at Reddit, Inc.

Seattle, Washington, United States
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

🤩
Rockstar
🎓
Top School
Phil Yeres is a software engineer with 11 years of experience building ML platforms and model-serving systems across major tech companies, currently driving ML platform work at Reddit. He has built production-focused AI infrastructure at TikTok and Apple and contributed research-grade systems for LLM fine-tuning and evaluation at NYU. His background includes hands-on work in model interpretability at NVIDIA and practical NLP toolkit improvements—adding ANLI tasks and gradient accumulation—to the widely used jiant repository. Based in Seattle, he blends research sensibilities with production engineering, shipping scalable training and serving pipelines. Colleagues describe him as someone who moves fluidly between low-level trainer refactors and large-scale deployment concerns, making models more reproducible and maintainable in real-world systems.
code11 years of coding experience
job3 years of employment as a software developer
bookNew York University
bookSwarthmore College
stackoverflow-logo

Stackoverflow

Stats
1reputation
0reached
0answers
0questions
github-logo-circle

Github Skills (9)

transformers10
pytorch10
nlp10
multi-task-learning10
python10
bert9
machine-learning9
data-preprocessing9
transfer-learning8

Programming languages (3)

CSSJupyter NotebookPython

Github contributions (5)

github-logo-circle
nyu-mll/jiant

Nov 2019 - Aug 2020

jiant is an nlp toolkit
Role in this project:
userML Engineer
Contributions:4 releases, 88 commits, 64 PRs in 8 months
Contributions summary:Phil primarily contributed to the development of the `jiant` NLP toolkit, adding new tasks related to Adversarial Natural Language Inference (ANLI) and improving the training process. They implemented gradient accumulation as an option for the trainer, enhancing the model training capabilities. Additionally, the user refactored and cleaned up code in the trainer and pre-processing modules, which would contribute to the overall maintainability and usability of the toolkit.
nlptransformersmultitask-learningsentence-representationbert
pyeres/pix2pix-pytorch

Apr 2017 - May 2017

Contributions:56 pushes in 9 days
pytorchdeep-learningadversarialconditionaltranslation
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