● Covered by 1 source · 1 reportLow impact1 neutral
Jumio Implements Real-Time Feature Store on AWS for Fraud Detection
Key points
- Jumio built a real-time feature store on AWS.
- The store addresses data duplication and latency for ML models.
- It uses SageMaker Feature Store, Flink, and Kinesis Data Streams.
- The system provides sub-100ms latency for predictions.
Addressing ML Challenges with Real-Time Features
Jumio, a provider of identity verification services, developed a real-time feature store on AWS to enhance its fraud detection capabilities. The company's machine learning models required immediate access to features to make accurate decisions, but faced issues such as data duplication, inconsistent feature definitions across teams, and manual deployment processes that introduced risks of mismatches and bugs.
Architecture and AWS Services Utilized
The new architecture leverages several AWS services to create a centralized, reusable, and real-time feature store. Key components include Amazon SageMaker Feature Store for managing features, Amazon Managed Service for Apache Flink for real-time data processing, and Amazon Kinesis Data Streams for data ingestion. This setup is designed to support machine learning use cases requiring sub-100ms latency for predictions.
Impact on Jumio's Operations
Before this implementation, feature engineering and deployment were fragmented and inefficient. The new real-time feature store aims to resolve these issues by providing a unified platform. This allows Jumio to maintain feature consistency, automate deployments, and reduce latency, which is critical for timely and accurate fraud detection in identity verification solutions.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Reporting from
Jumio, an identity verification provider, built a real-time feature store on AWS to address challenges like data duplication, inconsistent feature definitions, and latency in its fraud detection machine learning models. This implementation uses Amazon SageMaker Feature Store, Amazon Managed Service for Apache Flink, and Amazon Kinesis Data Streams to provide sub-100ms latency for real-time predictions.