• Standardized regression testing for the transit analytics page with Jenkins by replicating the production environment and creating procedures to test the application’s behavior and characteristics

• Wrote test scripts using Python to integrate all backend components (Kafka, Dataflow, Cloud Storage, and Elasticsearch) into a single program for simulating production-equivalent environment scenarios

• Expedited the cloud analytics feature launch into production by reporting over 17 bug tickets (out of 32 tickets) from May to August by following through with the bug fixes

AI tech studio | Kiwami Yosoku AI (Vertex Prediction AI)

· Implemented preprocessor, trainer and predictor API for estimating the quality of google responsive display ads based on delivery settings, description and images with TensorFlow, Docker, Kubernetes and GCP

· Developed automatic weekly model update system for delivering real-time score prediction which reflects recent advertising trends using Airflow

・Work closely with enterprise clients for integrating applications and automating business process on Workato's platform with various APIs