Led the data-tech team on data development and infrastructure R&D.
•Established real-time data processing service with Flink and Kafka, built the first real-time dashboards for sales events and daily business monitor.
•Applied tree-algorithm in batch data processing optimization, reduced data production time from 70 to 30 minutes, saved 20% computing & storage resources.
Designed and developed event-tracking management platform with Java/Vue.
•Developed real-time event-tracking validation service with Flink/Redis.
•Developed customized dashboard and toolbox for event-tracking data analysis, significantly improved efficiency compared to request BI to manually export data.
•Replaced 3rd party service, saved cost and controlled data leakage risk.
Rearchitected data query middleware to enable version/process control, circuit breaker, flow monitor and task group isolation, improved reliability above 99.99%.
Customized Metabase as a data analysis platform to replace 3rd party service.