Derrick Lin

⚬ Built an IP protection system for TikTok Shop to protect merchant IP rights, combining algorithmic and human moderation to process 200k reports per month.

⚬ Architected a product image protection system, integrating vector search and GPT-based similarity analysis on 80M images to detect unauthorized image reuse. ⚬ Built an extensible strategy configuration system supporting nested filters and adaptive queueing for Ops, cutting report routing strategies update time from 1 week to 1 day. ⚬ Designed an FaaS pipeline to resolve 20k problematic moderation tasks, enhancing user experience and significantly reducing the moderators’ backlog.

⚬ Enhanced caching resilience with feature toggles for digital banking services backend to increase system reliability. ⚬ Designed and implemented high-performance and OpenAPI-compliant APIs using asynchronous programming in Java Vert.x, serving 2000 debit transactions per hour.

Google Scholar: https://scholar.google.com/citations?user=yxhnw-0AAAAJ ⚬ Wireless Mobile Network Lab (Prof. Hung-Yu Wei) ⚬ Conducted researches in resource allocation optimization using edge computing and game theory. ⚬ First-authored conference paper "Collaborative Edge Caching with Multiple Virtual Reality Service Providers Using Coalition Games" accepted and presented in IEEE WCNC 2023. ⚬ First-authored journal paper “Multi-Service Edge Computing Management with Multi-Stage Coalition Game Task Offloading” published in IEEE TNSM Volume: 21, Issue: 3, June 2024.

⚬ Architected and implemented a scalable microservices architecture for serving 21 million users in Taiwan. ⚬ Designed and implemented a high-performance data retrieval and storage system, utilizing cronjobs for automation and multithreading for efficient transfer of 27M records from Amazon S3 to Redis in under 10 minutes. ⚬ Created CI pipelines building Docker images and measuring code quality with SonarQube, automated deployment to Kubernetes clusters with ArgoCD, and monitored with Grafana and Prometheus.

⚬ Trained an ML model to predict if an incident needs manual review or intervention. ⚬ Created a CI pipeline to periodically fetch incident data from ServiceNow, classify incidents using the trained ML model, and update to Google Sheets, reducing manual monitoring time by 30 minutes per day. ⚬ Presented to the US team and is used in production.

⚬ Built a unified platform integrating multiple ordering methods, payment systems, delivery services, and invoice systems, serving over 200 restaurant owners and 10k customers. ⚬ Developed full-stack features using jQuery and Laravel, with MySQL and MongoDB for storage. ⚬ Created a CI/CD pipeline to containerize the application and push the Docker image to Google Kubernetes Engine automatically.

GPA 3.9/4 Coursework: ⚬ CS 411 Database Systems ⚬ CS 425 Distributed Systems ⚬ CS 427 Software Engineering ⚬ CS 441 Applied Machine Learning ⚬ CS 463 Computer Security II ⚬ CS 513 Theory & Practice of Data Cleaning ⚬ CS 519 Scientific Visualization ⚬ CS 598 Foundations of Data Curation

Minor: Economics Labs: Wireless Mobile Network Lab, Network and Systems Lab Activities and societies: Captain of EE Department Football Team, Judo Varsity Team