Case Studies Hazelcast Powers Real-Time Fraud Detection

Edit This Case Study Record

	Hazelcast Powers Real-Time Fraud Detection

Hazelcast Powers Real-Time Fraud Detection

Technology Category	Analytics & Modeling - Real Time Analytics Application Infrastructure & Middleware - Data Exchange & Integration Application Infrastructure & Middleware - Database Management & Storage
Applicable Industries	Finance & Insurance
Applicable Functions	Business Operation
Use Cases	Fraud Detection Real-Time Location System (RTLS)
Services	Software Design & Engineering Services System Integration
Challenge	Competent fraud detection systems are crucial for organizations to identify financial crimes such as payment card fraud, anti-money laundering, and anti-terrorist financing. The increasing sophistication of fraudsters' systems and the growing volume of financial transactions put immense pressure on existing detection systems. In the UK, payment card spending exceeds £180 billion from 3.9 billion transactions, with financial fraud losses consistently rising. Detection systems must be flexible, report anomalies in near real-time, and be available 24/7 to cope with these challenges. Read More
About Customer	A top ten US bank faced significant challenges with its existing fraud detection system, which was built on an old relational database platform. The system's limitations caused the bank to hit transaction rate limits, breaking fraud detection SLAs and hindering new business opportunities. The bank needed a more efficient and scalable solution to handle its growing data and transaction volumes, ensuring real-time fraud detection and compliance with SLAs. Read More
Solution	Hazelcast In-Memory Data Grid (IMDG) and Hazelcast Jet were implemented to address the bank's challenges. Hazelcast IMDG stores account details, recent transaction history, and scoring in one place, enabling powerful and comprehensive checks within short SLAs of 500ms. The in-memory technology allows for fast data ingestion and transformation, while the distributed architecture ensures high availability and scalability. Hazelcast Jet provides in-memory stream and fast batch processing, further enhancing the system's real-time capabilities. The bank also uses WAN replication to synchronize multiple Hazelcast clusters, ensuring service continuity in case of regional data center outages. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Real Time Analytics

Application Infrastructure & Middleware - Data Exchange & Integration

Application Infrastructure & Middleware - Database Management & Storage

Applicable Industries

Finance & Insurance

Applicable Functions

Business Operation

Use Cases

Fraud Detection

Real-Time Location System (RTLS)

Services

Software Design & Engineering Services

System Integration

Challenge

Competent fraud detection systems are crucial for organizations to identify financial crimes such as payment card fraud, anti-money laundering, and anti-terrorist financing. The increasing sophistication of fraudsters' systems and the growing volume of financial transactions put immense pressure on existing detection systems. In the UK, payment card spending exceeds £180 billion from 3.9 billion transactions, with financial fraud losses consistently rising. Detection systems must be flexible, report anomalies in near real-time, and be available 24/7 to cope with these challenges.

About Customer

A top ten US bank faced significant challenges with its existing fraud detection system, which was built on an old relational database platform. The system's limitations caused the bank to hit transaction rate limits, breaking fraud detection SLAs and hindering new business opportunities. The bank needed a more efficient and scalable solution to handle its growing data and transaction volumes, ensuring real-time fraud detection and compliance with SLAs.

Solution

Hazelcast In-Memory Data Grid (IMDG) and Hazelcast Jet were implemented to address the bank's challenges. Hazelcast IMDG stores account details, recent transaction history, and scoring in one place, enabling powerful and comprehensive checks within short SLAs of 500ms. The in-memory technology allows for fast data ingestion and transformation, while the distributed architecture ensures high availability and scalability. Hazelcast Jet provides in-memory stream and fast batch processing, further enhancing the system's real-time capabilities. The bank also uses WAN replication to synchronize multiple Hazelcast clusters, ensuring service continuity in case of regional data center outages.

Impact #1	Hazelcast's in-memory technology allows the bank to process 5,000 transactions per second, with expectations to grow to 10,000 transactions.
Impact #2	The distributed architecture of Hazelcast ensures high availability and scalability, allowing the system to handle large volumes of data and transactions without performance degradation.
Impact #3	The implementation of Hazelcast has enabled the bank to meet its fraud detection SLAs, ensuring real-time detection and compliance.

Impact #1

Hazelcast's in-memory technology allows the bank to process 5,000 transactions per second, with expectations to grow to 10,000 transactions.

Impact #2

The distributed architecture of Hazelcast ensures high availability and scalability, allowing the system to handle large volumes of data and transactions without performance degradation.

Impact #3

The implementation of Hazelcast has enabled the bank to meet its fraud detection SLAs, ensuring real-time detection and compliance.

Benefit #1	Hazelcast is used to store 2TB of customer data, with plans to grow to 5TB.
Benefit #2	The bank can process 5,000 transactions per second, with expectations to grow to 10,000 transactions.

Benefit #1

Hazelcast is used to store 2TB of customer data, with plans to grow to 5TB.

Benefit #2

The bank can process 5,000 transactions per second, with expectations to grow to 10,000 transactions.

Download PDF Version

Overview

Hazelcast Powers Real-Time Fraud Detection

Operational Impact

Quantitative Benefit