What does this Amazon Web Services Solution do?
This solution provides a real-time anti-fraud system architecture based on deep learning graph neural network. The solution can be used in various scenarios, such as e-commerce websites, live video, games, and social applications. The solution detects abnormal business behaviors, such as fraud transactions, data crawling, and false loan applications.
The solution architecture is built with Amazon Neptune, Amazon SageMaker, Amazon Glue, etc. The core of AI is implemented using the Deep Graph Library (DGL) library, which is an open source package built for easy implementation of graph neural network model family, on top of deep learning frameworks. The deep learning technology behind the solution is backed by Amazon Web Services AI Lab, which aims to bring you the most cutting-edge technology.
The solution comes with an end-to-end demo which simulates the financial fraud detection. The demo uses the tabular data of the IEEE-CIS open dataset to construct a heterogeneous graph and trains the graph neural network(GNN) model for the real-time fraudulent transaction detection. From the demo dashboard, you can see the hundreds of thousands of online transactions are generated and the suspicious transactions are detected.
Amazon Web Services Solution overview
The below two diagrams show the “real-time fraud detection” architecture and “offline model training” architecture. Please contact to Amazon Web Services Sales to obtain the deployment guide and Amazon CloudFormation template.
The solution deploys the following resources:
Amazon Virtual Private Cloud (Amazon VPC) network topology contains the solution's resources.
Amazon Step Functions orchestrates the pipeline of training GNN model on DGL and deploys the SageMaker inference endpoint. The pipeline interacts with Amazon Simple Storage Service (Amazon S3), Amazon Neptune, Amazon Glue, Amazon Elastic Container Service (ECS) and Amazon SageMaker to process data, import graph data into graph database, train the model and deploy inference endpoint.
Amazon Lambda processes the real-time transaction request, then queries the graph database Amazon Neptune and inference the fraudulent possibility of transaction via Amazon SageMaker endpoint.
A web console powered by Amazon Amplify JS library is deployed into an Amazon S3 bucket configured for static web hosting. Amazon SQS, Amazon Lambda, Amazon AppSync and Amazon DocumentDB are used as backend services.
Amazon CloudFront provides secure, public access to the solution’s website content in the S3 bucket.
After you deploy this solution, you can train the machine learning model of the solution in Amazon Step Functions. You can open the demo web UI to simulate the transaction requests and observe the fraud transactions being detected in real-time.
Cutting-edge AI technology
Open Source and customization
Browse our portfolio of Amazon Web Services-built solutions to common architectural problems.
Find Amazon Web Services certified consulting and technology partners to help you get started.
Sign-up and start exploring our services.