I am a Machine Learning Engineer with experience in LLM/GenAI application development, and a focus on improving ML pipeline reliability and deploying scalable distributed systems. Skilled in Python, PyTorch, Spark, TypeScript and end-to-end data workflows, I focus on delivering production-oriented AI development and efficient data pipelines.

Yidan Wang

I am a Machine Learning Engineer with experience in LLM/GenAI application development, and a focus on improving ML pipeline reliability and deploying scalable distributed systems. Skilled in Python, PyTorch, Spark, TypeScript and end-to-end data workflows, I focus on delivering production-oriented AI development and efficient data pipelines.

Available to hire

I am a Machine Learning Engineer with experience in LLM/GenAI application development, and a focus on improving ML pipeline reliability and deploying scalable distributed systems.

Skilled in Python, PyTorch, Spark, TypeScript and end-to-end data workflows, I focus on delivering production-oriented AI development and efficient data pipelines.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Intermediate
Intermediate
Intermediate
See more

Language

English
Fluent

Work Experience

AI Engineer at VoiceAI Technologies Co. Ltd.
May 1, 2025 - August 24, 2025
Built automated data crawling and preprocessing scripts using BeautifulSoup and Flask, developed a customized stop-word dictionary and used NLTK for data cleaning to collect and structure large-scale domain-specific dialogue data. Developed and fine-tuned a transformer-based semantic similarity model using HuggingFace and PyTorch achieving a 12% precision improvement. Designed an emotion classification model combining Named Entity Recognition and rule-based weighting enabling 8-category emotion classification for context-aware agent replies. Implemented Redis-based caching to reduce model latency and consistently recognized for analytical thinking, efficiency, and clear communication.
Research Intern at University of Waterloo
August 1, 2024 - August 24, 2025
Developed a reusable TypeScript component with React to parse and render .ipynb files as editable HTML slides enabling one-click transformation of code notebooks to presentations. Implemented Markdown-to-HTML parsing pipeline with remark and rehype for lecture content rendering. Integrated RESTful APIs using Axios for dynamic user data interactions. Optimized performance with lazy loading and route-based rendering for large files, reducing load time and improving interactivity.
Team Lead & Frontend Developer at Android Task Management App Project
August 1, 2022 - August 24, 2025
Led the team developing a task management app supporting real-time location tagging, multi-status tracking, and cross-session data consistency to streamline student collaboration. Independently developed frontend in Kotlin, validated REST endpoints with Postman, and authored integration tests. Conducted UI/UX refinement through structured user interviews to optimize responsiveness and usability. Project was ranked first in final evaluation and recognized for technical leadership and effective collaboration.
Lead Developer at UniCore (Reddit-style Web Platform)
May 1, 2023 - August 24, 2025
Led development of UniCore, a university-focused peer discussion platform, designed UI/UX components in React.js, Next.js, TypeScript, Tailwind CSS, and Material UI. Implemented backend functionalities including authentication with JSON Web Token and Azure AD B2C for secure access. Validated OAuth2/OIDC flows and negative cases to ensure robust identity management. Applied lazy loading and route-based code splitting in Next.js to optimize performance and reduce load time.
Machine Learning Engineer at VoiceAI Technologies
August 1, 2024 - August 1, 2024
Optimized semantic-similarity calculation pipeline using Transformers (HuggingFace), redesigned preprocessing strategy (self-defined weighted keyword filter strategy, stopword dictionary), improving precision by 12% in deployed conversational flows. Architected a low-latency inference layer implementing Redis-based caching and dynamic batching, increasing system throughput under high concurrency and reducing average inference latency by 35%. Fine-tuned an 8-class emotion classification BERT-based model by integrating model outputs with self-designed entity-recognition–driven weighting rules, significantly stabilizing predictions for rare emotional categories by 20%. Delivered production-ready ML endpoints via FastAPI and Docker with clear interface contracts, enabling smooth integration by backend teams and reducing deployment issues.

Education

Master of Engineering in Electrical and Computer Engineering at University of Waterloo
September 1, 2023 - September 1, 2024
Bachelor of Engineering in Artificial Intelligence at Harbin Institute of Technology (HIT)
August 1, 2019 - June 1, 2023
Master of Engineering at University of Waterloo
September 1, 2023 - December 1, 2024
Bachelor of Engineering at Harbin Institute of Technology
August 1, 2019 - June 1, 2023

Qualifications

National Third Prize in China (International) Sensor Innovation Competition
January 1, 2023 - August 24, 2025
Excellent Student Leader and Excellent Volunteer
January 1, 2019 - December 31, 2023

Industry Experience

Software & Internet, Education, Computers & Electronics, Media & Entertainment, Professional Services