Case Study: HUMAN Security accelerates ML training and time to market with AWS Glue

HUMAN Security Accelerates ML Training and Time to Market Using Amazon SageMaker

HUMAN Security, a cybersecurity company, needed to speed up the manual training and deployment of its machine learning models for MediaGuard, its ad fraud detection solution. The company wanted to reduce weeks-long deployment cycles, improve model performance, and better handle growing data volumes, and it used AWS services including Amazon SageMaker and AWS Glue to support that effort.

AWS helped HUMAN Security automate data preparation, model training, offline evaluation, and deployment. With AWS Glue preparing data and Amazon SageMaker building, training, and deploying models, HUMAN Security cut training time from weeks to hours, tripled the number of ML models in production, ingested 5x more data, and now validates 15 trillion online interactions each week.

HUMAN Security

Austin Leirvik

Staff Data Scientist