Jiahao Sun
AI Engineer at OpusClip
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolJiahao Sun is an AI Engineer based in San Francisco with five years of hands-on experience building end-to-end ML systems, from ETL and feature engineering to LLM fine-tuning, deployment, and monitoring. He has driven measurable impact—improving retrieval relevance by 30% with an agentic RAG QA system, fine-tuning LLaMA3.1-8B to >90% precision/recall, and boosting model performance and trial enrollment through productionized monitoring and retraining on SageMaker. Comfortable across MLOps, scalable pipelines, causal inference and experimental design, he has shipped services with FastAPI, Docker, Airflow and hybrid search stacks (BM25 + embeddings + FAISS). An active contributor to OpenMMLab’s mmdetection3d, he optimized 3D detection performance via torch.permute refactors and voxel feature encoders—evidence of low-level efficiency tuning as well as high-level system design. His background blends applied data science (USC) and business analytics, enabling him to translate complex ML research into reliable, high-impact production solutions.
5 years of coding experience
4 years of employment as a software developer
Master of Science - MS, Applied Data Science, 3.9/4.0, Master of Science - MS, Applied Data Science, 3.9/4.0 at University of Southern CaliforniaMaster of Science - MS, Managment Science with specialization in Business Analytics, 3.89/4.0, Master of Science - MS, Managment Science with specialization in Business Analytics, 3.89/4.0 at Case Western Reserve UniversityBachelor of Science - BS, International Economics, 3.51/4.0, Bachelor of Science - BS, International Economics, 3.51/4.0 at Zhejiang University
English, Chinese
Github Skills (6)
object-detection10
pytorch10
3d-object-detection10
tensor10
tensorflow10
point-cloud10
Programming languages (2)
Github contributions (5)
open-mmlab/mmdetection3d
Jan 2023 - Jan 2023
OpenMMLab's next-generation platform for general 3D object detection.
Role in this project:
ML Engineer
ML EngineerContributions:4 releases, 97 reviews, 1 commit in 1 day
Contributions summary:Jiahao made several contributions focused on optimizing the performance and functionality of 3D object detection within the repository. They refactored code to utilize `torch.permute` for faster computation, implemented cylindrical voxelization and a voxel feature encoder, and added support for a torchsparse wrapper. The user also addressed bugs related to data formatting and testing, specifically focusing on the datasets and transformation pipelines.
pytorchnext-generationkitti-datasetdeep-learningobject-detection
sunjiahao1999/mmdetection3d
Jan 2023 - Apr 2024
OpenMMLab's next-generation platform for general 3D object detection.
Contributions:255 pushes, 63 branches in 1 year 3 months
pytorchnext-generationdeep-learningobject-detectioncomputer-vision
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.