AI Denied-Persons Screening
Global Trade-Compliance Platform
The brief
A trade-compliance software company had researched a deep-learning approach to natural-language classification and captured it in an academic whitepaper, backed by a simple proof-of-concept codebase. They needed that research turned into a production system.
What we built
We engineered the research into a scalable, cloud-native architecture on AWS. We then applied the lessons learned to rebuild the second version of their denied-persons screening tool, using SageMaker, Kinesis, Step Functions and S3 to move data through the pipeline at scale.
Outcome
The system became one of the cornerstones of the client’s platform.
