Network Behavior Analysis Project‣ Delivered the detection results for hourly malware detection on 20 million NetFlows per hour.‣ Saved 80% of team members’ time by developing a network anomaly detection engine as the main developer to help members to develop a new algorithm based on the engine.‣ Achieved 95% accuracy on the network anomaly detection engine with an ensemble algorithm to integrate all the algorithms in the engine.‣ Built a data pipeline solution to transform 600GB encrypted network traffic per day with my team via AWS Lambda and Glue (PySpark).Spam Website Detection Project‣ Accelerated the modeling development time by 500% by constructing complete CI/CD flow from GitHub PR through Jenkins to AWS Fargate for Kubernetes clusters of products.‣ Increased the time to update model for data scientists by 300% via assisting in writing unit tests with at least 80% coverage rate on testing cases.‣ Decreased the response time by 60% by designing and building a synchronized website crawler to collect needed features to input detection models.‣ Improved the detection rate of a NLP-based DNN model on the scam website detection by 140% via adding features on websites.‣ Led two junior software engineers to deliver the new feature of products in Trend Micro in 3 months with an improved detection rate from 2% to 45%.Consumer Renewal Data Analysis Project‣ Demonstrated success on two cross-discipline projects with the marketing team and the UI/UX team to increase company’s revenue.‣ Saved 70% of time of marketing team to deliver the new promotion event for stopping customers from leaving via CI/CD and AWS Fargate to predict the possible customer loss.