Kai T. Email & Phone Number

Sunnyvale, Ca, Us

Menlo Park, Ca, Us

Working on Facebook Ads Monetization Team to design and implement advertising optimization strategies.

Austin, Texas, Us

Worked on Underwriting Automation Team to build an automation platform to review a credit application for loaning Tesla vehicle.

San Francisco, California, Us

Container Management Tool (Infrastructure Development | Go)• Worked on the Compute Platform team to develop a container management tool for 9,800 images per repository on Amazon ECR.• Customized Amazon ECR lifecycle policies by Go and managed authorization on AWS IAM roles and policies by Terraform.• Streamlined the deployment of the tool for multiple regions to Kubernetes(Amazon EKS) by Helm chart, Spinnaker and Jenkins.• Achieved the automatic synchronous management of containers in different regions, reducing more than 70% training time.

Developed a real-time video platform on Tammira App for learners and instructors using JavaScript and Agora, ReactNative, Node.js, and Amazon Simple Storage Service(S3).

San Francisco, California, Us

Worked on an open-source command-line utility and API to support single-command application deployment and service catalogs (chart repositories) on Nomad.

Merchant-side APP (Cross-Platform Mobile Development | JavaScript & Java)• Worked with two engineers and a product manager to build a personal dine-in assistant Android and IOS App for merchants to view and confirm orders, manage reservations, update and recommend menus, and give refunds by React Native and JaveScript.• Designed dynamic and interactive UI with smooth animations and gestures by React-Navigation.• Developed RESTful APIs by Swagger(API description tool) for visualizing APIs operations and automate test case generation.• Accomplished the app platform migration from an Android to both Android and IOS Apps.

Wa, Seattle, Us

A Dynamic Model for Predicting Bus-arrival Time (Big Data | Python) KCMT is the eighth-largest transit bus agency in the US. https://github.com/tank606/KCM-Capstone (made by Figma)• Worked with six engineers and two data scientists to improve customer experience on the current transit system by incorporating a predictive model by Python and Azure. • Used filter-based features selection method to analyze top factors that influence bus arrival time and dynamic time warping forclustering and supervised learning models (Boosted decision tree, neural network, etc.) for prediction.• Implemented this machine learning pipeline using Azure ML Studio and tested it on the test dataset of 300 stations.

Amsterdam, Noord-Holland, Nl

Medical Imaging Analysis Software (Windows Software Development | C++) Collaborated with three engineers to design a computer-aided diagnosis tool to assist doctors with GUI component written using Qt.Extracted morphometry features from medical images (Python, C++, OpenCV, TensorFlow), stored (Navicat, MySQL), analyzed (Cloud AutoML Google) and visualized data (RStudio) to evaluate brain blood vessel conditions.Effective in reducing manual review time by 40% for both experts and novice radiologists in clinical reading.

Master'S Degree, Information And Communication Engineering

Doctor Of Philosophy - Phd, Electrical And Computer Engineering