Skip to main content

AWS Solutions Library

Guidance for Detecting Point of Sale Fraud using Amazon SageMaker

Overview

This Guidance shows how your travel and hospitality (T&H) business can effectively detect fraudulent transactions and gain fraud prediction insights on AWS. Amazon SageMaker processes point-of-sale (POS) terminal data and detects anomalies, then uses ML to predict fraud and quickly alert fraud prevention teams. Additionally, Amazon Bedrock converts fraud prediction datasets for natural language querying so that your leadership team can gain understanding through a chatbot UI by using plain English. By using this Guidance, you can better detect and prevent fraud in real time while garnering insights to support decisions that protect your business assets.

How it works

This architecture diagram shows how T&H businesses can effectively detect fraudulent transactions at POS terminals and quickly gain insights from fraud prediction datasets by using Amazon SageMaker and Amazon Bedrock.

Download the architecture diagram

Well-Architected Pillars

The architecture diagram above is an example of a Solution created with Well-Architected best practices in mind. To be fully Well-Architected, you should follow as many Well-Architected best practices as possible.

SageMaker simplifies building, training, and deploying ML models, and QuickSight enables you to create interactive dashboards and visualizations without specialized skills or complex data engineering. As fully managed services, SageMaker and QuickSight handle infrastructure scaling, maintenance, and automation. This frees you to focus on optimizing ML models and gaining insights from data visualizations, rather than managing underlying complexities.

Read the Operational Excellence whitepaper

This Guidance uses AWS Identity and Access Management (IAM)to support least privilege access and role separation, reducing the risk of unauthorized access or actions. Additionally, Amazon S3 bucket policies support bucket access control, encryption at rest, version and object locking, and access logging. These policies enable you to make sure that only authorized roles, users, or services can read or write data to the buckets. You can also use the log data for security analysis, auditing, and compliance purposes. Finally, QuickSight supports row-level security, enabling you to control access to specific data rows based on user or group membership, preventing unauthorized data exposure.

Read the Security whitepaper

SageMaker provides automatic scaling, fault tolerance, and model validation features to support reliable and accurate ML model deployment. It automatically distributes training data and models across multiple Availability Zones (AZs) and enables automated retraining to maintain model accuracy over time. Additionally, Amazon S3 provides reliable and durable data storage capabilities through its built-in high-availability features. For example, it replicates data across multiple AZs and redundant storage facilities, minimizing data loss and supporting consistent data access.

Read the Reliability whitepaper

SageMaker offers optimized hardware instances, distributed training, and automatic model tuning to improve the performance and efficiency of your ML workloads. For example, you can choose optimal instance types, and your workload scales based on demand, resulting in efficient utilization of compute and memory resources. Additionally, Amazon S3 provides highly scalable and low-latency data storage, enabling efficient data ingestion and retrieval for training and deploying ML models.

Read the Performance Efficiency whitepaper

Amazon S3 offers lifecycle policies and storage classes that optimize storage and costs. For example, Amazon S3 Intelligent-Tiering automatically moves objects to the most cost-effective storage tier based on access patterns. Additionally, SageMaker offers per-second billing for training instances, so you only pay for the resources you use, and it scales automatically to optimize costs. And as a managed service, it removes the overhead of provisioning and managing your own ML infrastructure.

Read the Cost Optimization whitepaper

Amazon Bedrock provides a sustainable foundation for running data-intensive workloads. By scaling automatically, it optimizes resource usage and avoids the environmental impact of overprovisioned or inefficient infrastructure. Additionally, Amazon S3 provides highly durable and secure object storage that protects data integrity, minimizing the need for redundant backups. It also supports efficient data storage and management practices, such as Amazon S3 Lifecycle policies like Amazon S3 Intelligent-Tiering and object expiration. Using these management capabilities, you can minimize the storage resources required for your workloads and reduce the energy consumption associated with storing and processing data.

Read the Sustainability whitepaper

Disclaimer

The sample code; software libraries; command line tools; proofs of concept; templates; or other related technology (including any of the foregoing that are provided by our personnel) is provided to you as AWS Content under the AWS Customer Agreement, or the relevant written agreement between you and AWS (whichever applies). You should not use this AWS Content in your production accounts, or on production or other critical data. You are responsible for testing, securing, and optimizing the AWS Content, such as sample code, as appropriate for production grade use based on your specific quality control practices and standards. Deploying AWS Content may incur AWS charges for creating or using AWS chargeable resources, such as running Amazon EC2 instances or using Amazon S3 storage.