Wen Tat Woong

- Main PIC of antifraud data warehouse, continuously serving business by fulfilling data requests, increasing data warehouse coverage and adoption rate, and providing training and support to BI users - Maintained SLA baseline for antifraud business line critical company reports - Developed graph data warehouse for antifraud, providing entity linkage data and clustering results in online and offline tools - Tech stack: Java, Python, Jupyter, Shell, Golang, React, Hive, Hudi, Hbase, Clickhouse, Kafka, Flink, Spark, Trino

- Set up 10 PB+ data warehouse in LATAM IDC to serve LATAM BI and ops in a timely manner, developed and maintained cross IDC pipelines between SEA and LATAM data warehouses - Migrated our data warehouse processing from Hive to HUDI to utilize incremental processing, speeding up batch jobs from daily to hourly - Migrated our biggest kafka topic of 500k+ QPS from JSON to AVRO to save on ETL cost and kafka bandwidth, reducing message size by 50\% - Utilized clustering techniques to compress Parquet files by 50\%, saving up to 10PB+ of storage - Automated downstream Hive tables schema and ETL changes based on schema registry, to support upstream services' weekly regular release - Developed Golang backend service to sample 20000+ Kafka topics for sensitive data auditing, discovered and fixed 1000+ data vulnerabilities

- Developed and designed antifraud data warehouse of 30PB+, utilizing dimensional modelling and data warehouse layering - Developed batch jobs handling TB level of data, and stream jobs handling 500k+ QPS - Developed pipeline to ingest 50k+ QPS into Clickhouse and Golang backend service serving fraud logs with REST and GraphQL APIs

Backend engineer for central ads database: - Build data pipeline for ads data service using Hive, Azkaban, Java streaming jobs - Set up monitoring and log processing infrastructure with ELK stack and Grafana - Benchmarked several noSQL databases with YCSB for new database migration - Maintained and managed critical infrastructure such as Couchbase, Kafka, ELK stack and Grafana

- Online entertainment company developing a fantasy league platform to increase engagement of esports with esports audience. - Spearheaded product development efforts, from creating Google Forms prototype to developing a Flask + React website under 3 months. - Designed and developed database schema and domain specific language for defining fantasy leagues. - Automated data collection by writing web scrapers. - Set up user interviews to gather information and prioritize feature requests. - Set up user analytics with Google Analytics and Mixpanel to inform design decisions. - Marketed products in physical events such as Dota2 KL Major and PvP esports championship.

- Worked on A.I powered fraud detection system codebase. - Aggregated data for dashboard, reducing data redundancy and speeding up load times by 3x - Automated business intelligence efforts to review fraudulent data by creating internal tools, eliminating 5+ man hours per week.