I work on buidling LLM infrastructure as a Staff ML System Engineer at OctoAI (now NVIDIA). I received my Ph.D. in Computer Science at Carnegie Mellon University with a dissertation titled "Automated and Portable Machine Learning Systems". At CMU, I was fortunate to be advised by Prof. Tianqi Chen and Prof. Zhihao Jia as a member of the Automated ML System group. I am passionate about co-designing and co-optimizing algorithm and system for efficient ML, concentrating on multi-modal and large language models. To this end, I have worked towards building automated and portable system for distributed ML with a focus on parallelism, operator fusion, and graph optimization.

Email: byungsoj.com [at] gmail.com | | | | |

Teaching

CMU 15-884 Machine Learning Systems

Teaching Assitant, Spring 2021, Instructors: Tianqi Chen

CMU 10-403 Deep Reinforcement Learning and Control

Teaching Assitant, Spring 2020, Instructors: Katerina Fragkiadaki

CMU 10-701 Machine Learning (PhD)

Teaching Assitant, Spring 2019, Instructors: Leila Wehbe, Aaditya Ramdas

Award

Qualcomm Innovation Fellowship 2022

Winner in North America, awarded $100K to the team (Byungsoo Jeon, Sunghyun Kim from MIT)
- Project title: Holistic Distributed Deep Learning Compilation with Automated Cross-stack Optimization

Kwanjeong Scholarship 2017 - 2021

Awarded from Kwanjeong Educational Foundation for promising PhD students. ($30,000 per year)

Publication

> Automated and Portable ML System

GraphPipe: Improving the Performance and Scalability of DNN Training with Graph Pipeline Parallelism

Byungsoo Jeon*, Mengdi Wu*, Sunghyun Kim*, Shiyi Cao*, Sunghyun Park, Neeraj Aggarwal, Colin Unger, Daiyaan Arfeen, Peiyuan Liao, Xupeng Miao, Mohammad Alizadeh, Gregory R. Ganger, Tianqi Chen, Zhihao Jia
ASPLOS 2025
PDF

Collage: Seamless Integration of Deep Learning Backends with Automatic Placement

Byungsoo Jeon*, Sunghyun Park*, Peiyuan Liao, Sheng Xu, Tianqi Chen, Zhihao Jia
PACT 2022 - Integrated to Apache TVM Project (in TVM v0.9.0) / Presented in GTC 2022
PDF | Slides | Code

SRTuner: Effective Compiler Optimization Customization by Exposing Synergistic Relations

Sunghyun Park, Salar Latifi, Yongjun Park, Armand Behroozi, Byungsoo Jeon, Scott Mahlke
CGO 2022
PDF

> Applied ML / RL

OBP-RL: Exploring Deep Reinforcement Learning Methods for Online Binpacking Problem

Byungsoo Jeon, Bharathan Balaji, Saurabh Gupta, Chun Ye
Amazon Machine Leanring Conference 2020

FactoredRL: Leveraging factored graphs for deep reinforcement learning

Bharathan Balaji*, Petros Christodoulou*, Xiaoyu lu*, Byungsoo Jeon, Jordan Bell-Masterson
NeurIPS 2020 (Deep RL Workshop)
PDF

Dropout Prediction over Weeks in MOOCs by Learning Representations of Clicks and Videos

Byungsoo Jeon*, Namyong Park*
AAAI 2020 (AI4Edu Workshop)
PDF

Dropout Prediction over Weeks in MOOCs via Interpretable Multi-Layer Representation Learning

Byungsoo Jeon*, Namyong Park*, Seojin Bang*
AAAI 2020 (AI4Edu Workshop)
PDF

Time-series Insights into the Process of Passing or Failing Online University Courses using Neural-Induced Interpretable Student States

Byungsoo Jeon, Eyal Shafran, Luke Breitfeller, Jason Levin, Carolyn P. Rose
EDM 2019 (Short oral presentation)
PDF

Attentive Interaction Model: Modeling Changes in View in Argumentation

Yohan Jo, Shivani Poddar, Byungsoo Jeon, Qinlan Shen, Carolyn P. Rose, Graham Neubig
NAACL 2018 (Poster)
PDF | Code

Music Emotion Recognition via End-to-End Multimodal Neural Networks

Byungsoo Jeon, Chanju Kim, Adrian Kim, Dongwon Kim, Jangyeon Park, Jungwoo Ha,
RecSys 2017 (Poster)
PDF