Year2016 – Ongoing
RoleDevOps, Big Data Stack, Machine Learning
TechnologyHDFS, ElasticSearch, Postgres, Spark, Kafka, NiFi, plus custom dev in Java, Scala, and Node.JS
Fight financial crime using AI / ML
A client dealing in financial crime prediction platform were faced with processing around eight billion transactions a month. They wanted a high performance and flexible AML alerting system with Machine Learning techniques to handle the large volume of data. Further, they also desired to significantly reduce false positive rates for alerts.
Sense, predict and optimize AML process
We brought in a suite of tools to integrate with internal systems such as legacy data sources. And built an always-on stream processing system with the Big Data stack. It was configurable to the client infrastructure—public or private cloud—and optimized end-to-end digital security management.
AML screening with reduced false positive alerts
The AML solution was able to identify suspicious activities through Machine Learning capabilities and provide richer insights on potential AML risk exposures including pictures, connections, and media mentions. False positive alerts were reduced as well.