*Developed data pipeline's stages, including building segregation, tagging test accounts, and creating notification event stage for streamlined data management

* Integrated Metadata Registry into data pipeline, enabling schema retrieval and designing fetch metadata stage to support varied attribute metadata fetching

* Engineered and implemented List of Map merge logic for database operations, ensuring seamless data integration and storage efficiency

* Investigated metrics in Flink to transmit essential data such as error counts, event throughput, and latency to Wavefront dashboard; utilized Wavefront API to replicate dashboard templates across 300 pipelines

* Implemented dropFFA, segregation, and consent logic (7216) within data pipeline to ensure events are correctly routed to designated buckets for notifications

* Enhanced existing read event process to trigger notifications post-database storage, significantly reducing read costs on DynamoDB in subsequent stages

* Refined pipeline logic by incorporating additional properties early in the execution process, minimizing redundant operations and enhancing efficiency

* Took charge of AWS EMR restacking, individually managed PageDuty alerts, and leveraged Splunk for debugging pipeline issues

* Designed copilot alert module to find advertising campaign error, such as creative rejection and no fire beacon, and send the email to notify user every day

* Restyled console UI, such as campaign targeting section, campaign default section and bid forecaster, and designed the new campaign health dashboard

* Implemented concept stacking targeting frontend and backend allow user create concept targeting in workflow

* Designed programs to count distinct active users by every minute and hour in real time using Java

* Created data visualization system to present different services’ active users in real time on web using Java Spring framework, JavaScript and bootstrap