I am an Engineer in Amazon Prime Video Personalization and Discovery, focused on recommendation systems and personalization. My daily work covers distributed backend services, ranking pipelines, online model serving, offline parameters tuning, batch ML inference, and production operations.

I care about feed recommendation quality in the same spirit as LinkedIn/TikTok-style products: stable high-throughput serving, strong ranking relevance, and fast offline-to-online iteration with clear metric guardrails.

Before Amazon, I worked at Baidu USA on applied AI systems and earlier in autonomous driving research. I earned an M.S. at the University of Michigan and an M.S. at CUHK. This background shaped how I build recommendation systems: system-first, data-driven, and production-oriented.

Experience

Oct 2024 – Present
Software Engineer, Machine Learning
Prime Video Personalization and Discovery (PVPD), Seattle, WA
Recommendation platform engineering: ranking integration, AWS CDK infrastructure, distributed cache for low-latency serving, EMR-based batch inference, and monitoring/alerting guardrails.
Jan 2024 – Jul 2024
Research Engineer Intern
Baidu USA, Sunnyvale, CA
Text-to-video generation (Kubrick), autonomous cement truck deployment on Apollo platform.
Sep 2022 – Dec 2023
Graduate Research Assistant
University of Michigan
Drone Remote ID systems, privacy analysis, urban air mobility under Prof. Max Z. Li.
Aug 2020 – Aug 2022
Research Associate
The Chinese University of Hong Kong (CUHK), Hong Kong
Control theory, SOS programming, barrier functions, RL for safety-critical systems under Dr. Dongkun Han.
Mar 2019 – Jul 2019
Intellectual Property Intern
Daimler Greater China, Beijing

Habits & Beyond Code

Weekend Mode

Fishing

Lake mornings help me think slower and cleaner. Fish are honest reviewers: no bite means the strategy is wrong.

Team Mode

Football and Soccer

I started following football during my time in Ann Arbor, and I've been a huge fan of Lionel Messi ever since.

Creative Mode

Music

Saxophone and live bands keep me grounded. Good rhythm helps in code too: fewer surprises, better flow, cleaner handoff.

Education

M.S. Aerospace Engineering

University of Michigan, Ann Arbor

2022 – 2023

M.S. Mechanical and Automation Engineering

The Chinese University of Hong Kong

2020 – 2022

B.Eng. Mechatronics Engineering

North China Electric Power University

2015 – 2019

Technical Skills

Languages

Python Kotlin Typescript Java C++ MATLAB

Recommendation Modeling

Ranking Pipelines Re-ranking Strategy Feature Engineering Offline Evaluation A/B Test Experimentation Auto Tuning

ML Infrastructure

AWS CDK AWS SageMaker Bedrock Spark PySpark EMR Cluster EMR Serverless Ray SGLang DynamoDB

Serving & Inference

Online Model Serving Batch ML Inference Distributed Cache Monitoring Alerting Java Services Python Pipelines Docker Kubernetes Heap tuning

Academic Service

Conference Reviewer: CVPR, NeurIPS, ICML, MIDL, CCTA
Teaching Assistant: ENGG1910 Demystifying AI, CUHK, 2022
Summer Research Project, CUHK, 2021 & 2022