Worked on the design and development of low latency Microservices to cater the needs of PayPal partners/merchants. These are low-latency APIs and are developed using a mix of technologies including but not limited to Java, Spring Boot, Google Cloud Bigtable (NoSQL), Juno (PayPal's Key-Value store) and the APIs are deployed on k8s cluster with high availability.Contributed to the development and maintenance of a reporting platform to cater the reporting needs of PayPal customers. This framework supports multiple sources and multiple destinations along with pre-defined Data Quality checks. The platform is generic in terms that new sources and destinations can be supported easily. Adding new Data Quality checks or Custom User Defined Functions can be done with ease. It is built using Apache Spark, Scala and runs on the Google Cloud Dataproc cluster.Currently working on the design and development of Unified Transactions product which will provide a unified view of the transactions happening across PayPal.

As a part of the experimentation team, I worked on building a Streaming Data Platform which could create real time metrics for User Telemetry and Transaction Telemetry.

Built multi source, multi sink platform using Apache Spark which is used in MLaaS framework for doing Auto ML.Migrated SQL Server procedures of 1000+ lines to Hive scripts.Helped the Data Science team in architecting and building the necessary infrastructure so that they could use Pyspark for building their Machine Learning models.

Built Data Quality Framework from scratch which could be used to find correctness of the data, based on which, Machine Learning team could make better decisions for building their ML models.