At Socure, I developed device and behavioral biometrics products to detect fraud. I created gRPC services in Java and TypeScript for behavioral biometrics, stored in DynamoDB. Additionally, I built RESTful services in Scala to identify risky devices using DynamoDB, Elastic Search Cluster, and Redshift data warehousing.
Moreover, I prototyped a pipeline using AWS Kinesis Data Stream, Firehose, Glue Jobs, and Elastic Search Cluster for generating device risk scores. Finally, I managed an ETL pipeline in Scala, integrating SQS queues and Redshift data warehousing, providing our data science team with extensive device data for analysis.
Tech Stack: Scala, Java, TypeScript, RESTful, gRPC, SQS queue, DynamoDB, Elastic Search Cluster, Redshift data warehouse, AWS kinesis data stream, firehose, glue job, Spring boot, Scalatra, Akka, Google Guice.
Funded by the National Science Foundation (NSF), the goal of the research was to replace super-conducting magnets with powered magnets in Magnetic Resonance (MR) imaging machines to make them smaller, affordable, and cryogen-free. I designed computer simulations to filter MR images from field fluctuations found in powered magnets. Furthermore, I collaborated with researchers from Penn State University, Florida State University, and National High Magnetic Field Lab, Florida to share findings.
