국내 No.1 에너지 IT기업 ‘해줌’의 컨플루언트 클라우드 도입 스토리 | 알아보고 등록하기

Featured Partner: PT. Mitra Integrasi Informatika (MII)

Fraud Detection System for Banking Customer

In the banking sector, traditional fraud detection methods often struggle with time-consuming manual reviews, scalability issues, delayed detection, and high costs. Garda addresses these challenges by leveraging Confluent's real-time data streaming capabilities.

Partner use case - Build with Confluent badge

Protect Businesses and Customers from Fraudulent Activities



Garda is a highly scalable Fraud Detection System developed by PT. Mitra Integrasi Informatika (MII) that has capabilities that include advanced analytics, machine learning, real-time monitoring, alerting, and behavioral analysis.

Garda uses customized connectors to collect data from different sources and uses stream processing and rule processing. Processed data is made available to Elasticsearch and ML job processing to power the Garda dashboard.

By incorporating these capabilities, the Garda real-time fraud detection system can effectively protect businesses and their customers from fraudulent activities, ensuring a secure and trustworthy environment.

Benefits of the solution include:

Improved Efficiency

Garda automates the fraud detection process, significantly reducing the time and resources required for manual reviews.

Real-Time Detection

By processing data in real-time, Garda can detect and prevent fraudulent activities before they cause harm.

Enhanced Security

Garda's advanced analytics and machine learning capabilities provide a robust defense against sophisticated fraud schemes.

Reduced Costs

By automating fraud detection, Garda can help organizations reduce operational costs associated with manual review processes.

Build with Confluent

This use case leverages the following building blocks in Confluent Cloud:

Reference Architecture



Connect to Data Sources

Allowing efficient data ingestion from multiple sources, including databases, card transactions, and digital channels, ensuring a comprehensive view of activities.

Stream Processing

Leveraging Confluent to enable flow of data in real time, and using Flink as a stream processing engine to implement rules on data anomalies classified as fraud.

Machine Learning

Using a sink connector to connect Elastic, which functions as a machine learning engine on Garda.

Garda Dashboard

Monitoring data identified as fraud by Garda and also configuring the rules that will be applied to the fraud detection system.

Resources

Contact Us

Contact MII to learn more about this use case and get started with Garda.