About

Yung-En Lu

Master of Computer Science @ Rice University Aug 2025 – Dec 2026

Master of Science in Computer Science @ National Yang Ming Chiao Tung University Sep 2024 – Jun 2025

Bachelor of Science in Computer Science and Information Engineering @ National Cheng Kung University Sep 2021 – Jun 2024

Software Engineer Intern @ MaiAgent (formerly Playma) Feb 2025 – Present

Developed new features with Django and automated DevOps workflows for the MaiAgent AI platform
Reduced CNAME vanity domain setup time by over 80% for our AI webchats by developing a backend service that automates CDN provisioning and SSL/TLS certificate management on AWS CloudFront
Reduced production deployment time for 12 AWS EC2 instances from 30 minutes to under 5 minutes by utilizing Ansible
Enabled on-premises deployment for 3 new enterprise customers by creating Ansible playbooks to automate setup across 5 diverse environments, including air-gapped networks(Ubuntu/RHEL)
Integrated HeyGen and D-ID services into the webchat to create talking avatars for customer engagement
Setup CI pipelines using GitHub Actions, automating unit and integration testing
Reduced Ansible test duration by 50% across diverse inventory setups by implementing CI pipelines with Molecule and GitHub Actions
Designed and implemented CI pipelines using GitHub Actions, automating unit and integration testing, which caught 20% of production errors during pre-deployment testing
Setup CI pipelines for unit and integration testing and securtiy for our on-premise Django application with Trivy, Sonarqube and testing for Ansible playbooks

Software Engineer Intern @ Trend Micro Jul 2023 – Aug 2023

Worked on the Web Reputation Service team, contributing to the migration of the Web Classify Server (WCS), a legacy service handling 700 million daily classification requests
Cut infrastructure costs by 10% by migrating the WCS legacy service from VMs to Kubernetes
Reduced WCS deployment time by 20% by building CI/CD pipelines with GitHub Actions and Helm charts
Designed and implemented a Prometheus metrics exporter for WCS, enabling real-time error detection and reducing reliance on user-reported errors
Reduced desynchronization errors with AWS S3 Cross-Region Replication by designing and implementing the WCS pattern file upload process

Software Engineer @ AInimal Jan 2022 – Jun 2023

Led backend migration from monolith to microservices, reducing the development cycle by 40%
Led development of the gRPC API for backend microservices, increasing backend system fault tolerance
Increased maximum concurrent users to over 5,000 by adopting Golang and NoSQL databases
Wrote unit tests, integration tests and load tests for chat service with over 90% code coverage. Set up GitLab CI/CD to automate testing and building container images
Redesigned the user notification feature and optimized the click-through process, reducing interaction time by 20% and enhancing user experience

Software Engineer Intern @ Intelligent Mobile Service Lab Jun 2022 – Jan 2023

Developed a responsive online judge website using Vue and Vuetify, enhancing the user experience for over 200 students
Improved page load times and responsiveness by refactoring core components and implementing asynchronous component loading

Research Assistant @ Software Engineering and Intelligent Test Automation Lab Oct 2021 – Oct 2024

Developed a prompt optimization method for large language models (LLMs) using pairwise feedback, reducing annotation effort by 50% while maintaining performance
Implemented a new screen recording feature for Rapi (web application testing software), enabling local and remote browser recording during test cases which reduced debugging and bug reproduction time by 50%
Built an auto-scalable tool for a Kubernetes-based Selenium Grid, enabling reliable video capture across large-scale parallel test suites

Remote Research Assistant @ XR for Intelligent Medicine Lab Dec 2023 – Sep 2024

Developed a customizable LLM-based virtual therapist using LangChain, retrieval-augmented generation (RAG), and prompt engineering, enhancing personalized mental health support
Surveyed and experimented with different evaluation methods to assess the effectiveness of an LLM-based virtual therapist in providing emotional support

DBonK8s | Go, Kubernetes, Terraform

A chatbot that gives users and admins an easy way to manage databases for development and testing

Utilized Kubernetes API, namespace and role-based access control to manage resources, users, and administrators
Supports four different types of databases (MySQL, PostgreSQL, Redis, MongoDB) with custom settings
Utilized Terraform to deploy to Google Cloud Platform and Kustomize to manage manifest files
Built a chatbot to manage four types of databases (MySQL, PostgreSQL, Redis, MongoDB), giving developers and admins an easy way to manage resources for development and testing
Utilized Terraform to deploy to Google Cloud Platform and Kustomize to manage manifest files
Developed a chatbot to streamline database management (MySQL, PostgreSQL, Redis, MongoDB), simplifying resource provisioning for development and testing environments
Built a chatbot to simplify database provisioning (MySQL, PostgreSQL, Redis, MongoDB) for development and testing environments
Configured Kubernetes role-based access control (RBAC) and namespaces to create secure, multi-tenant resource management
Utilized Kubernetes API, namespace and role-based access control (RBAC) to create secure, multi-tenant resource management
Utilized Terraform and Kustomize for provisioning and configuration management

Yungen’s blog | Hugo, GitHub Actions

My personal website that automatically syncs with my Obsidian notes through GitHub Actions

Developed a personal website that automatically syncs with Obsidian notes
Implemented a CI/CD pipeline using GitHub Actions for content sync and Cloudflare Pages for deployment bookmark | TypeScript, Node.js, Hugo, GitHub Actions

My personal bookmark website that automatically syncs with my Notion notes through GitHub Actions

Utilized Notion SDK to export Markdown files from Notion notes and Hugo to generate a static site
Automated the generation, updating, and deployment processes using GitHub Actions, eliminating the need for manual reconfiguration and deployment, resulting in a 30-minute time savings per edit

Yung-En Lu and Shin-Jie Lee, “Prompt Optimization with Human Annotations Using Pairwise Comparison,” in Proceedings of the IEEE International Conference on e-Business Engineering (ICEBE), 2024, pp. 15–22, doi: 10.1109/ICEBE62490.2024.00012.
Chun-Chuan Chen, Meng-Chang Tsai, Eric Hsiao-Kuang Wu, Shao-Rong Sheng, Jia-Jeng Lee, Yung-En Lu, Shih-Ching Yeh, “Fusion model using resting neurophysiological data to help mass screening of methamphetamine use disorder,” IEEE Journal of Translational Engineering in Health and Medicine, doi: 10.1109/JTEHM.2024.3522356.