Résumé

I build and operate CI/CD pipelines, containerized microservices, and cloud-native MLOps systems. Looking for a role where I can own systems end-to-end.

Software Engineer 🔗MLOps Engineer 🔗iOS Developer 🔗Extended 🔗

01 ／Education

New York UniversityMay 2026
New York, NY · GPA 3.73 · Computer Engineering, M.S.
Practical Software EngineeringHuman Computer InteractionMachine LearningMLOpsReinforcement Learning
National Taichung University of Science and Technology (NTCUST)June 2023
Taichung, Taiwan · GPA 3.79 · Computer Science and Information Engineering, B.Eng.
Software EngineeringElectronic Commerce SecurityAlgorithmsData Structures

02 ／Technical Skills

Languages
PythonGoJavaC#SwiftJavaScriptSQL (PostgreSQL, MS SQL)C/C++
Cloud & DevOps
DockerAWSTerraformGitHub ActionsAirflowPrometheusGrafanaLinuxGit/GitFlow
Frameworks & Backend
FastAPIFlaskDjangoNode.jsReactVueRESTful APIsWebSocket
Data & ML
PyTorchMLflowLlamaIndexLangChainLightGBMSHAPStreamlit

03 ／Professional Experience

6 roles
Co-Founder at A DayMay 2025 – Present
New York, NY
  • Architected a production-grade MVVM application with SwiftData and latest SwiftUI features, decoupling the service layer to ensure strict separation of concerns across 7 interconnected data models and 60+ source files.
  • Engineered an AI-augmented development workflow (Cursor/Claude) with custom MCPs and skills for context management, accelerating delivery 4× while maintaining TDD cycles to catch regressions.
  • Established a CI/CD pipeline on Xcode Cloud with Swift Testing, implementing 700+ unit and integration tests across ViewModels and data flows.
Course Assistant at New York UniversityJanuary 2026 – May 2026
Brooklyn, NY · Part-time
  • Built a 5-microservice ML safeguard lab (FastAPI, Docker Compose) routing requests through two custom MobileNetV2 guard models and a Food11 classifier via short-circuit logic; generated synthetic edge-case data and benchmarked median/p95/p99 inference latency to quantify the cost of each safety layer.
  • Reviewed end-to-end MLOps systems from 30 teams (110 students) on Chameleon Cloud across design, code, and live demos, assessing them against production criteria: experiment tracking, automated retraining, drift monitoring, secrets hygiene, and container reproducibility; delivered structured feedback that teams used to remediate issues such as data leakage and shortcut/proxy learning.
Data Science Intern at Micron TechnologyJuly 2025 – August 2025
Taoyuan, Taiwan · Internship
  • Architected a production-scale Python pipeline and self-serve Streamlit app over 33GB of bi-weekly fab-dispatch logs, surfacing key scheduling drivers via an explainable LightGBM/SHAP model for reproducible, evidence-based tuning adopted across teams.
  • Engineered repo-documentation tools for an enterprise application with over one million lines of code using Prompt Engineering and sub-task agents; cut token usage 10x and projected ~3x gain in developer efficiency.
Software Developer at CARITY AIMay 2024 – August 2024
Ontario, Canada · Full-time
  • Automated CI/CD for an LLM-based product by containerizing 4 microservices on AWS ECS with GitHub Actions, reducing infrastructure costs by 40% and cutting deployment time by 70%.
  • Delivered a Retrieval-Augmented Generation (RAG) proof-of-concept showing a potential 5x token-usage reduction, shaping the team's cost-optimization roadmap.
Software Engineer at MoBagelJanuary 2023 – July 2023
Taichung, Taiwan · Full-time
  • Engineered a full-stack material management system for the Taiwan Water Corporation (national utility, 7M+ households), migrating its legacy Java application to automate inventory and budgeting over millions of historical records.
  • Introduced GitFlow and an Agile workflow that scaled the engineering process as the team tripled from 10 to 30, enabling a smooth transition recognized by the supervising manager.
  • Discovered and reported critical security flaws, including SQL-injection vectors and a publicly exposed, unauthenticated test database that risked a large-scale breach of decades of records.
Software Engineering Intern at Mindtronic AIJune 2022 – September 2022
Taipei, Taiwan · Internship
  • Rebuilt the entire backend in Go from a legacy Node.js service, owning 53 RESTful APIs end to end (design, implementation, and SQL-injection hardening) while reliably processing 480K+ entries weekly.

04 ／Projects

9 selected
NYU MarketplaceSeptember 2025 – December 2025
Practical Software Engineering, NYU · New York, NY
  • Architected a CI/CD pipeline on Travis CI and AWS Elastic Beanstalk with strict merge gates (linting, 85%+ test coverage), automating deployment on every approved PR.
  • Built RESTful and real-time WebSocket APIs with Django 5 and Channels, and reviewed all team PRs for system design and code quality.
  • Advised by Professor Gennadiy Civil (Adjunct; Senior Engineering Manager at Google).
Privacy‑First AI Smart Lamp for Ephemeral Night Conversations - OblivilightJuly 2025 – August 2025
OpenHCI'25 · Taipei, Taiwan
  • Led a user-centric design process from research to prototype, identifying key user needs for tangible, privacy-preserving "forgetting mechanisms" in AI companions through 11 user interviews and secondary research.
  • Architected a full-stack proof-of-concept integrating an LLM for conversation, emotion analysis, and a multi-modal interface with voice (Whisper/TTS) and gesture controls.
  • Designed an interaction model that visualizes emotional sentiment as colored light and externalizes digital conversations into physical artifacts via a thermal printer, addressing AI data permanence anxiety.
  • Won the Best Demo Award at OpenHCI'25 and was the sole project selected for presentation at the subsequent TAICHI'25 conference.
Real‑Time Plant Health & Mood Visualization - Loud Plants in Your AreaFebruary 2025 – May 2025
NYU · New York, NY
  • Led the development of an iOS application that translates plant bio-acoustic signals into real-time AR visualizations, defining the project vision and architecting the full stack.
  • Engineered a custom machine learning pipeline based on academic research to classify plant health. Independently implemented a deep scattering network (ScatNet with Morlet wavelets), extracted Mel-frequency cepstral coefficients (MFCCs), and trained a high-performing SVM classifier for signal analysis.
  • Developed a fully functional AR prototype using Swift, RealityKit, and Reality Composer Pro. Owned the entire iOS application development, building custom animated UI overlays that rendered dynamic plant statuses based on live data from the ML pipeline.
  • Validated the project's core hypothesis through a user evaluation study that demonstrated the AR interface significantly increased user-plant interaction and emotional connection (mean score increase from 1.67 to 5.33, p=0.018).
Taigi (Taiwanese-Hokkien) Medical Advising LLMMarch 2025 – May 2025
NYU · New York, NY
  • Architected a cloud-native MLOps platform for LLM using Terraform for IaC, and deployed a suite of Docker-based microservices (FastAPI, Gradio, MinIO) to production.
  • Orchestrated a Continuous Training (CT) pipeline with Airflow for human-in-the-loop retraining, and established system observability using Prometheus and Grafana.
  • Fine-tuned an 8B LLaMA-3.1 into the first Taigi medical advisor using 120K bilingual Q&A pairs with LoRA + mixed-precision on an A100 GPU; tracked all runs in MLflow for full reproducibility.
RISC-V-SimulatorOctober 2024 – December 2024
NYU · New York, NY
  • Architected a RISC-V processor simulator in Python, implementing both single-stage and five-stage pipelined designs to analyze and compare architectural performance trade-offs.
  • Engineered the hazard management system for the five-stage pipeline, with data forwarding units, a load-use hazard detection unit for stall insertion, and a branch forwarding unit with early resolution in the ID stage to minimize control hazards.
  • Developed a parallel simulation workflow to execute identical RISC-V programs on both pipelines simultaneously, ensuring accurate, side-by-side comparison of outputs and performance metrics.
  • Built an automated performance analysis module that calculates and reports key metrics, including total cycle count, instructions per cycle (IPC), and cycles per instruction (CPI) for each processor core.
AI Editor-in-Chief and Virtual News PresenterSeptember 2021 – July 2023
Natural Language Process Lab, NTCUST · Taichung, Taiwan
  • Led a year-long capstone project from concept to award-winning completion, architecting a fully automated AI pipeline that autonomously generated animated news segments from trending topics.
  • Engineered the core system infrastructure to resolve dependency and versioning conflicts across 5 open-source microservices; designed a data pipeline using Docker Compose and Flask for reliable integration.
  • Automated the end-to-end deployment process for the entire stack, creating a reproducible, one-command build that slashed manual setup and deployment time by over 80% (from 2 hours to 20 minutes).
  • Introduced GitFlow to the team, establishing a structured version control workflow that sped up development and collaboration, which I later applied at MoBagel during a 3× team expansion.
TW-COVID-Bot (Telegram Bot)August 2020 – November 2022
Spontaneous Study · Taipei, Taiwan
  • During the COVID-19 pandemic in Taiwan, I assisted social media influencers in automatically posting daily outbreak information for easy access to up-to-date information.
  • Volunteered to organize developers to set up a server and developed a bot that utilized web crawling, image generating, and data analysis.
  • Automatically graphing daily COVID-19 cases from 200 countries to track the level of spread.
RetailpiaSeptember 2021
Level-Up Society Hackathon · UK
  • Led a team of 4 students from 4 universities to develop a city-building game with an account book app and points collector within 48 hours.
  • Winner of Better Retail, Level-Up Society Hackathon, organized by ShowCode.
iOS DeveloperJune 2020 – October 2020
I.M.A.C. Lab, NTCUST · Taichung, Taiwan
  • Designed a weather app with MVVM, implementing weather API, Realm Database, and MapKit.
  • Designed a bank app with MVC, implementing biometric authenticating, notification broadcasting, and real-time currency displaying.
  • Utilized Swift UI layout systems such as Storyboards and SwiftUI with Swift5.

05 ／Honors & Awards

1st prize & Best Demo2025
OpenHCI’25, presented at the 11th Annual TAICHI Conference · Taipei, Taiwan
Emerging Technology Application Award2023
Fi-Award 2023 by the 13th International Conference on Frontier Computing · Tokyo, Japan
1st prize in Project Competition2023
College of Information and Distribution Science, NTCUST · Taichung, Taiwan
Winner of Better Retail2021
Level-Up Society Hackathon · UK