November 2021 - Present

Software Engineer (Scala)  Agoda

• Agoda Intelligence (AI): Developed an AI system (10,000 hotels joined, 500k booking per day) designed for automated price adjustment. Utilized Kafka, Spark, MS SQL, and Scala technologies, and automated the internal data flow (calculating room night target) by building scheduled Spark jobs.

• Gitlab Migration: Migrated Agoda Intelligence from Github to Gitlab, and automated testing, static check and service mesh deployment in Gitlab CI pipeline, where developers can easily deploy AI and Spark jobs by just a few clicks on the Gitlab UI.

• K8s: Built  service cluster in K8s for AI, in which developers can deploy AI and its dependent services in K8s for testing within 1 minute.

• B2B Dynamic Price Adjustment: Developed and improved the Agoda B2B platform (millions USD transaction per month), a tool used by chain hotels for managing their marketing campaigns. Also, designed and implemented a dynamic pricing logic from scratch in backend, enabling hotels to effortlessly adjust their price margins via the B2B platform.

• Monitoring: Developed an e2e flow monitoring dashboard by leveraging Hadoop, Superset, and Airflow, enhancing the quality assurance of dynamic pricing adjustment in Agoda's B2B platform, which product manager, and developers can quickly and easily to monitor our product instead of running multiple SQL scripts by their own.

• Mentoring: Took the initiative to mentor and have knowledge sharing to team mates, fostering their professional growth and enriching the team culture. Recently, I organized a study group which focuses on the book, “Designing Data-Intensive Applications (DDIA).”

• On-call: Served as a 24-hour on-call engineer for Agoda's vital backend services occasionally. Successfully resolved a critical issue of Agoda core backend services to prevent potential financial loss from 10 thousands bookings.

December 2020 - November 2021

Sr. Software Engineer (Golang)  香港商慧科訊業有限公司

• API Design: Coordinated with teams based in Beijing, Taipei, and Shenzhen, we developed and implemented a high-standard API (gin-gonic), this cross-functional collaboration involved the Frontend, DevOps, Quality Assurance, and Data divisions.

• Leadership: Serving in the capacity of Technical Scrum Master, effectively managed the collaboration between the Front-End, Backend, and Data teams to strategically enhance our product, which currently serves a clientele comprising Fortune 500 corporations.

December 2019 - November 2020

Data Engineer  ETtoday 東森新聞雲

• gRPC: Undertook a comprehensive restructuring of our proprietary user behavior tracking system, integrating gRPC alongside the existing HTTP API for enhanced service performance.

• Ads Tracking: Formulated and implemented a sophisticated user behavior tracking mechanism, deeply integrated within ETtoday's news portal, to drive value-oriented advertisement placements.

• System Monitoring: Developed and instituted an anomaly detection system, leveraging the capabilities of Prometheus, to maintain robust surveillance over the data team's suite of services.

• Recommendation System: Implemented a dynamic and robust recommendation engine, powered by technologies such as Spark CF ALS, Redis, Kafka, and Cassandra, employed to enhance user experience on ETtoday's e-commerce platform.

• Graph Algorithm: Implemented the graph algorithm (trillion-level computing job on Spark in ~5 mins) aiming at the Entity Resolution problem.
• Microservices: Enhanced the mechanism of service discovery by adopting HashiCorp Consul.
• CICD: Enhanced the CD flow by introducing the Jenkins pipeline (Jenkinsfile) into the deployment workflow.
• ETL: Enhanced the front-end, integration test, and Spark UDF of ETtoday's in-house Spark-based ETL tool.

February 2019 - November 2019

Software Developer  NXP Semiconductors

• Mining the latent impact factors from the data of sales orders by collaborating with NXP's sales, operation, and finance team.
• Improved the development quality by defining the critical path test for NXP’s core business logic.
• Enhanced the data quality by refactoring the legacy Salesforce Apex code used in the approval flow of the sales business pipeline.