Joshua Yan
Software Engineer at Yutori
Victoria, British Columbia, Canada
Summary
🤩
Rockstar🎓
Top SchoolJoshua Yan is a software engineer from Victoria, BC with five years of experience building back-end infrastructure and ML systems, currently joining Jane Street and working on RL infrastructure and agent automation at Yutori. He has a strong academic foundation from the University of Waterloo (4.0 GPA) and practical ML experience in roles spanning inference acceleration, pre-training, and persistent kernels for draft models. Joshua contributes to prominent open-source tooling for LLMs—improving Ollama’s CLI, installation flow, model parsing, and template systems—demonstrating attention to both robustness and user-facing UX. His work blends research-minded rigor with production engineering: shipping infrastructure for sharing and inferencing large language models and accelerating logistic automation. Outside of code, he’s an avid Avalon player and jigsaw puzzler, a hint at his enjoyment of strategic play and complex problem decomposition.
5 years of coding experience
1 year of employment as a software developer
Mount Douglas Secondary SchoolBachelor's degree, Computer Science, 4.0 (94.3%), Bachelor's degree, Computer Science, 4.0 (94.3%) at University of Waterloo
Github Skills (5)
command-line-interface10
go10
command-line10
golang9
error-handling9
Programming languages (3)
Github contributions (5)
ollama/ollama
Feb 2024 - Aug 2024
Get up and running with Llama 3.3, DeepSeek-R1, Phi-4, Gemma 3, and other large language models.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:25 reviews, 56 PRs, 190 pushes in 6 months
Contributions summary:Joshua primarily contributed to the back-end of the Ollama project. Their work involved fixing word-wrapping issues and double-width character display in the command-line interface, indicating a focus on the user experience of interacting with language model outputs. They also made changes to the installation script, update URLs, and the model's error messages to improve the overall stability and user experience of the application. Furthermore, they made changes to the model parsing, and added a functionality related to the model's template system.
deepseekgemmagemma2gemma3go
joshyan1/arceus
Nov 2024 - Dec 2024
Build neural networks with Macbook clusters
Contributions:4 reviews, 30 PRs, 59 pushes in 15 days
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.