Joy Bhalla

Worked for a USA-based banking client, in the Market Risk division. As part of the backend team of a newly created project, was responsible for developing the ETL process and data pipeline. To source data from various internal banking systems, used Kaast and Apache Spark to extract the data and load it into HDFS. Leveraging Apache Spark using Scala, designed and implemented Spark jobs to read the data from HDFS, transform and load into required SQL database for end users. As part of moving to production, worked with Apace Oozie and the internal Python framework to schedule, and automate jobs for verifying data extraction and data load. As a result of our work, we were able to move from a monthly release to a biweekly release, moving from manually verifying data extraction and data load to an automated system.

Implemented Super Resolution in General Adversarial Networks. Implemented the model using tensorflow in the backend and keras in the front end. Deployed of model and training of data done one Floydhub.