Xinyuan Zhu

Xinyuan Zhu

AI & ML Engineer | Software Development

Yale University MS in Computer Science with a B.Eng from CUHK-Shenzhen. I build intelligent systems, machine learning, and real-world software products.

About Me

I am a driven computer science professional focused on building innovative solutions through AI and software engineering. My academic journey spans Yale University and The Chinese University of Hong Kong, Shenzhen.

My research interests include machine learning, deep learning, reinforcement learning, and robotics. I have hands-on experience from internships at Bosch, OPPO, and CITIC Securities.

Contact me directly at xinyuanzhu2024@163.com, or connect via LinkedIn and Instagram.

Phone: +86 18042788597

Email: xinyuanzhu2024@163.com

Location: China / USA

Education

Yale University

M.S.

Computer Science

Aug 2025 - Jul 2026

Pursuing advanced coursework in machine learning and artificial intelligence.

The Chinese University of Hong Kong, Shenzhen

B.Eng.

Computer Engineering

Jul 2021 - Jul 2025

  • GPA: 3.87 / 4.00 (1/78)
  • IELTS: 7.5
  • Top 1% academic honors
  • Outstanding Student Award

Technical University of Munich

Exchange

Computer Science & Engineering

Mar 2024 - Sep 2024

Exchange program with a focus on AI.

Experience

Bosch China

AI LLM Intern May 2025 - Jul 2025
  • Led workflow improvements for code deployment and test generation in a large language model project.
  • Optimized Bosch internal models using LoRA fine-tuning, raising coverage from 79% to 91%.
  • Supported 300+ daily model calls for engineering teams.
  • Built an end-to-end semantic understanding system with QLoRA-tuned local models.
  • Reduced development cycle time by over 30%.

OPPO Guangdong Mobile Communications

Software Development Intern Aug 2024 - Oct 2024
  • Built an automated progress alert bot that captured 100% of abnormal project events.
  • Contributed to the launch of the new OPPO App Store version.
  • Improved Android UI and feature flows across multiple releases.
  • Validated 23 test cycles and prevented four critical P0 defects.

CITIC Securities

Machine Learning Intern Feb 2024 - Mar 2024
  • Modeled CSI 300 index performance with XGBoost and technical indicators.
  • Achieved 70%+ prediction accuracy with feature engineering.
  • Enhanced model explainability using SHAP values.

Hong Kong Applied Science and Technology Research Institute

Machine Learning Intern Aug 2023 - Sep 2023
  • Developed a hybrid polynomial regression and random forest model.
  • Delivered a battery RUL prediction solution for sensor datasets.
  • Processed 100,000+ sensor records for reliability analysis.

Research

CUHK-Shenzhen Capstone Project

Sep 2024 - May 2025

  • Built a CNN-driven audio processing model for bowed instrument performance analysis.
  • Combined MFCC and spectrogram features to achieve 92% recognition accuracy.
  • Designed a multimodal architecture with Mel-spectrogram and time-domain features.
  • Incorporated dilated convolution and attention for long-sequence modeling.

Leibniz Supercomputing Centre at TUM

Apr 2024 - Aug 2024

  • Worked with a reinforcement learning team to reduce sim-to-real gap by 57%.
  • Tweaked Actor-Critic models for efficient transfer to physical environments.
  • Refined adaptive PID control strategies for smoother real-world behavior.

Shenzhen AI & Robotics Research Institute

Oct 2022 - Mar 2024

  • Worked as an undergraduate research assistant on Kalman filtering for vessel tracking.
  • Developed an adaptive algorithm for sailboat environment perception.
  • Contributed to published research in autonomous navigation.

Publications

[1]

CNNs in Musical Performance and Arrangement: Recognizing and Managing Bowed Instrument Techniques Across Cultures

X. Y. Zhu, C. Leung (2025)

2025 AIMEDIA, Venice, Italy

[2]

Design and Implementation of a Novel Adaptive Multihull Sailboat with Liftable Side Hulls

X. Y. Zhu[1], C. Liang[2], H. H. Qian* (2023)

2023 IEEE International Conference on Robotics and Biomimetics, Koh Samui, Thailand

Skills

Machine Learning

XGBoost Scikit-learn Decision Tree PCA SVM Classification & Regression K-Means

Databases & Tools

SQL Arduino IDE SolidWorks Android Studio Pandas NumPy Matplotlib Data Visualization

Artificial Intelligence

NLP (Fine-tuning, LLaMA, RAG, Agent) TensorFlow PyTorch Keras Deep Learning (CNN, Transformer, LSTM) Computer Vision Reinforcement Learning

Programming Languages

Python C/C++ C Java Verilog CUDA MATLAB

MLOps / DevOps

Docker Kubernetes Flask GitHub Actions

Frontend

HTML5 CSS3 JavaScript Kotlin

Contact

📞 Phone

+86 18042788597

🌍 Social

| |