ABOUT

This is an experimental AI human that can occasionally make mistakes.
Always verify important information.

Portrait of Junshen Kevin Chen

Hi

I'm generally interested in building AI things.

Over time I've found myself having worked on every part of the stack in an AI system, from training the model, to optimization, to inference algorithms, to large scale inference systems, to model orchestration, to the apps and the user interaction itself.

This website is for experimenting, ranting, memeing.

Experience

This is probably what you were looking for.

  1. JUL 2025 – PRESENT·LLM HARNESS AND NEW AI EXPERIENCES

    Meta Superintelligence Lab

  2. OCT 2023 – JUN 2025·DIFFUSION MODELS

    Meta Gen AI

  3. OCT 2021 – SEP 2023·COMPUTER VISION

    Meta Computational Photography

  4. 2018 – 2020·MASTER'S IN AI/ML

    Stanford University

  5. 2014 – 2018·B.S. COMPUTER SCIENCE

    UC San Diego

Resume

The most boring version possible.

&