Built distributed systems for ads from scratch as part of founding team (processing CPAs, clicks and impressions with Flink, Kafka, & Cassandra)

Designed and developed surfacing real time Ads data to merchants (with Apache Pinot, Kafka, and ETL scripts for historical data)

Developed features for DataVisor's fraud detection system (with Spark Streaming, Kafka, and Cassandra)

Implement deep learning solutions for question-answering, identifying relevant search candidates, and sequencing knowledge in knowledge bases.