(Working Remotely)Course Enrollment SystemI am involved as a full-stack developer in the system that serves 1500+ students.• Designed and created 5 tables in ​AWS RDS​ ​MySQL ​to allow teachers add/delete courses, students enroll/drop courses• Programmed enrollment website with ​JavaScript​ via ​React​ using ​React Router​, ​React Hook​, andMaterial UI​. Deployed on ​AWS S3 ​with static web hosting• Built a back-end service using ​Java​-based ​Spring Boot​. Published docker image with ​Google jib​ to ​AWS Elastic Container Registry (ECR)• Created and deployed dev & prod environment with 5 nodes each on ​AWS​ ​Elastic Kubernetes Service(EKS) ​with ​Elastic Load Balancer ​integrated• Introduced ​GitHub Action​ as ​Continuous Integration​ to reduce deployment time by 83%• Configured and connected MySQL with ​Hibernate​. Implemented all back-end APIs with ​RESTful convention. Introduced​ JWT token​ using ​Spring Security​ and stored it as cookies in front-endOnline Comment Stream Processor• Set u​p a data ingestion layer with internal live comment streaming API with ​Python• Implemented a real-time comment processor with ​Apache ​Spark​ ​for live-streaming comments• Introduced statistic ​chart for top 10 comments using ​Ajax​ long polling with 3s interval• Deployed and configured ​Google Cloud Platform VM instance​ to support external viewingReal Time Sentiment Analysis• Built event queuing layer with ​​Kafka​​ to process input dataflow• Implemented sentiment monitor with​ ​Apache Spark​​ capable of handling up to 60 comments per second• Visualized real-time positive vs. negative sentiments trends using ​Python• Created analytic dashboard using ​React ​to showcase different teachers’ popularity among online audience• Helped advertisement team to analyze audience’s preference, which increased course income by 10% and new user registration by 30%

Low Power Environmental Sensing System(Researcher under Prof. Bob Iannucci, CMU)• Commissioned by United State Geological Survey (USGS) with million dollars of funding• Developed a low-power wireless sensing system for collecting stream height data of Pepperwood Preserve, Sonoma County, California, achieved 10X battery life time extension with advance low power technology• Enabled lower power communication between sensing systems and gateway via LoRa • Reduced the cost of 5+ rangers/year for monitoring the stream height manually• Led a 4-engineer firmware team for this complicated system with 14 components• Implemented firmware for 5 sensors with hardware interfaces​ UART​,​ I2C​ and ​SPI ​using​ C• Used ​JTAG​, ​Logic Analyzer​ and​ Oscilloscope​ extensively• Used ​Python​ to orchestrate transmitter and receiver for LoRa signal strength data collection

Researcher under Prof. Vijay Sivaraman, UNSW • Managed a team of 4 engineering students to build a system that estimates the bus queue length, waiting time, and recommend the bus stop with shortest waiting time to users via mobile application• Utilized Python and mySQL API to implement user and bus data acquisition, resulting in a backend flask server bridging between web/mobile application and databases for 5 real-time parameters• Innovated the method of QR code scanning along the bus queue to track the queue length

• Enhanced over 10 existing BMW automatic car diagnosing system• Collaborated with 6 cross-functional teammates to refine over 20 Electronic Control Units circuit design• Communicated between over 6 customers and technical teams to help customers understand the technical transformation that their cars are going through in a daily basis

UNSWOT is a start-up program co-run by engineering and law students to build an online “Hackathon & Share” platform for 3 Sydney universities: UNSW, USYD and UTS• Tripled the user database from 60 to over 200 users in the first stage• Received an award for the most innovative ideas on the topic, Haze and Nuclear Power, and presented to over 50 professors and industry representatives