Zhen Tong
Machine Learning and Algorithm Engineer
About Me
I'm Zhen Tong, a machine learning and systems engineer with experience across deep learning, backend infrastructure, reinforcement learning, and generative modeling. My recent work spans large-scale cloud systems, research-oriented model building, and practical end-to-end engineering for AI applications.
Work Experience
Software Engineer at Google
Sunnyvale, California, United States | Hybrid | Feb 2026 - Present
Tech Stack: C++, gRPC
Google Cloud Dataflow core service infrastructure and job lifecycle management.
- Contributed to the core service infrastructure of Google Cloud Dataflow, improving system performance.
- Developed and improved job lifecycle management systems, with a focus on job status tracking and orchestration.
Software Engineer Intern at Amazon
Bellevue, Washington, United States | On-site | May 2025 - Aug 2025
Tech Stack: AWS, Java, Step Functions, Lambda
Built a centralized data deletion orchestration service for secure, large-scale customer data removal workflows.
- Designed and implemented a centralized deletion orchestration service that enabled Fulfillment teams to automate secure data removal in response to customer deletion requests.
- Reduced onboarding complexity across three internal AWS services by replacing custom pipelines with a standardized integration framework.
- Scaled throughput by 15x during multi-service onboarding by leveraging AWS Step Functions parallelism and stateless compute with AWS Lambda.
- Improved system resilience with idempotency, automatic retries, and partial deletion fallback to preserve consistency under failure scenarios.
Game AI Algorithm Intern at NetEase
Hangzhou, Zhejiang, China | On-site | May 2024 - Aug 2024
Tech Stack: Python, PyTorch, Ray, ETL
Reinforcement learning and imitation learning for poker-game AI systems.
- Designed efficient game state and action representations for a series of poker games, enabling effective reinforcement learning.
- Developed a distributed ETL pipeline using Ray to prepare large-scale game data in parquet format for model training.
- Implemented a deep imitation learning algorithm that achieved an 80% win rate against human experts.
- Deployed the trained agent and conducted extensive client-side performance testing, including concurrency tests.
Undergraduate Research Assistant at Learning of Graph & Optimization Lab
Shenzhen, Guangdong, China | On-site | Jan 2024 - May 2024
Tech Stack: Python, PyTorch
Proposed a novel benchmark and baseline model with Double-Stage Fusion for lane-level HD map generation.
- Proposed a dataset spanning 3-4 kilometers of road segments with incomplete lane lines to address challenges in cost-effective, real-world lane-level HD map generation.
- Designed a Double-Stage Fusion model that first uses a VQGAN to reconstruct input images, then trains a Transformer to learn the latent conditional distribution for generation.
- Evaluated the model on the EcoMap dataset, showing significant improvements over existing lane-level HD map generation methods.
- Presented the benchmark and baseline model at NeurIPS 2024.
Software Engineer Intern at Siemens
Foshan, Guangdong, China | On-site | Jul 2022 - Sep 2022
Tech Stack: Java, Hadoop, Python, PyTorch, MySQL
Built a deep learning anomaly diagnosis pipeline for a metro-control big data system and integrated the results into a visualization workflow.
- Selected useful data with MySQL, organized the relevant sources, and aligned them into a new real-time dataset.
- Trained an LSTM-RNN model in PyTorch to predict future system states, achieving 92% accuracy.
- Devised and deployed Hadoop-based infrastructure and efficient execution logic to scale prediction across large data volumes.
- Built a Java WebSocket backend with Maven to serve visualization requests from the front end.
Education
Master of Science in Information Networking @ Carnegie Mellon University
2024 - present
2024 Graduate Admission Scholarship
Bachelor of Computer Science @ Chinese University of Hong Kong ShenZhen
2020 - 2024
CGPA = 3.8 (rank 9)
2020 CUHKSZ School of Data Science Annual Scholarship
2020, 2021, 2022 CUHKSZ School of Data Science Dean List
UC Berkeley Global Access
2023 January - June
cs182 Designing, Visualizing and Understanding Deep Neural Networks Grade A
cs170 Efficient Algorithms and Intractable Problems Grade A
cs161 Computer Security Grade A-
GPA = 3.9
Teaching Experience
CUHKSZ
2021 - present
2021-2022 Summer CSC3100 Data Structure
2021-2022 Term2 CSC1002 Computational Laboratory of Python