I started this blog in late 2022 as a CS undergrad who wanted a place to put LeetCode solutions and course notes. It was built with NextJS, partly because I wanted to learn the framework and partly because I was tired of the MERN stack's SEO headaches.
A lot has changed since then. I've worked as a full-stack developer, a firmware engineer, an AI infrastructure engineer, and most recently as a researcher building RL training systems for LLMs at Huawei. I'm now heading into an MSc at the University of Alberta to work on reinforcement learning.
The blog has evolved too. What started as LeetCode dailies and tmux cheat sheets now includes write-ups on training-inference discrepancy, low-precision quantization on Ascend NPUs, and whatever else I'm deep in at the moment.
I don't post on a schedule — I write when I've figured something out and think it's worth explaining. If you're into RL, LLM infrastructure, or just like reading about engineers debugging tensor shapes at 11pm, you might find something useful here.