Full-Stack Engineer
TIM OH
Building at the intersection of AI + Full-stack engineering
STATUS: @ WATERLOO · SEEKING INTERNSHIPS
01
ADVANCED_MICRO_DEVICES_(AMD).EXE
AI Software Engineer Intern
Jan 2026 — Apr 2026Advanced Micro Devices (AMD)
- >Architected an end-to-end agentic AI triage system in Python that automatically discovers, ranks, and drafts GPU driver bug tickets from online user feedback, showcased to leadership and now being adopted org-wide.
- >Scaled issue triage to 10,000+ community posts per run by engineering a Python pipeline that combined semantic embeddings, LLM classification, and clustering to surface high-priority GPU issues.
- >Built a dual-pipeline Python architecture that reused shared collectors, LLM clients, and PostgreSQL/Redis infrastructure across both a triage workflow and MCP tool, eliminating duplicated retrieval and triage logic.
- >Orchestrated full-cycle Jira ticket drafting through an LLM agent chain that calls internal bug-tracking and crash-analysis APIs over MCP, with schema-validated outputs and source-linked evidence.
ADVANCED_MICRO_DEVICES_(AMD).EXE
Full-Stack Software Engineer Intern
Sept 2025 — Dec 2025Advanced Micro Devices (AMD)
- >Shipped a real-time AI reporting tool in React, Python, and FastAPI that captured live spoken feedback during GPU test runs, replacing manual note-taking with LLM-refined reports.
- >Engineered a low-latency speech-to-text pipeline with Whisper, Silero VAD, and PyTorch, tuning custom audio preprocessing and internal terminology biasing to achieve 93% accuracy on GPU-specific vocabulary.
- >Enhanced internal full-stack platforms built with React, Node.js, Express.js, Redux, and Snowflake, and shipped features for scheduling, real-time monitoring, and reporting across distributed GPU test plans.
FORD_MOTOR_COMPANY.EXE
Software Developer Intern
May 2022 — Aug 2022Ford Motor Company
- >Developed Python validation pipelines and reusable test harnesses for over-the-air firmware updates across Wi-Fi and cellular conditions, supporting release readiness and 99.9% update reliability across 5 connected-vehicle models.
- >Implemented CI and pre-release regression workflows by building a Python test-prioritization scheduler that ranked cases by historical failure rates, improving triage speed and reproducibility and saving 10+ developer hours per week.
- >Improved in-vehicle wireless stability by 15% by authoring Python regression suites for Wi-Fi hotspot and emergency-call features, root-causing handoff latency from multi-gigabyte packet captures.
02
03
SKILLS.EXE
// Languages
PythonTypeScriptJavaScriptC++CJavaSQLBashHTML/CSS
// Frameworks & Libraries
FastAPIReactNode.jsExpressDjangoFastMCPReduxTailwind CSS
// AI/ML
LLMsAgentic AIMCPRAGPrompt EngineeringEmbeddingsWhisperPyTorchLangChainLangGraph
// Databases & Tools
PostgreSQLRedisSnowflakeMongoDBDockerKubernetesAWSJenkinsGitLinux
04
EDUCATION.DAT
University of Waterloo
Bachelor of Mathematics, Co-op
Expected Dec 2027
President's Scholarship of Distinction
Extracurricular: Poker Studies Club, Computer Science Club, Data Science Club
05