Fraud Detection Using Machine Learning is an AWS Solution that automates the detection of potentially fraudulent activity, and flags that activity for review. The solution is easy to deploy and contains an example dataset. You can also modify the solution to use your own dataset. The update improves model accuracy and now includes a model to detect anomalies in unlabeled data. 

Now you can include enriched metadata in Amazon Virtual Private Cloud (Amazon VPC) flow logs published to Amazon CloudWatch Logs or Amazon Simple Storage Service (S3). Prior to this launch, custom format VPC flow logs enriched with additional metadata could be published only to S3. With this launch, we are also adding metadata fields that provide insights about the location of the network interface on which flow logs are being captured, such as the AWS Region, AWS Availability Zone, AWS Local Zone, or AWS Outpost where it resides.

AWS RoboMaker, a service that makes it easy to simulate and deploy robotics applications at cloud scale, now supports account-level metrics for simulation resources through Amazon CloudWatch. This feature helps customers more easily manage their resource limits to ensure that their simulation workloads continue to run successfully.  

Amazon Elasticsearch Service UltraWarm is now generally available. UltraWarm is a performance-optimized warm storage tier that lets you store and interactively analyze your data using Elasticsearch and Kibana while reducing your cost per GB by up to 90% over hot storage options. With UltraWarm, Amazon Elasticsearch Service supports hot-warm domain configurations. Hot storage is used for indexing and providing the fastest access to data. UltraWarm complements hot storage with less expensive, more durable storage for older data that you access less frequently, all while maintaining the same interactive analysis experience.

AWS Step Functions is now integrated with AWS CodeBuild, making it faster to build continuous integration workflows for your applications. Using AWS CodeBuild integration, you can easily create workflows with primitives such as branching, parallel execution, and timeouts to handle software changes from different repository branches. AWS Step Functions supports built-in error handling, parameter passing, recommended security settings , and state management, reducing the amount of code you have to write and maintain.

Amazon RDS Performance Insights supports SQL-level metrics on Amazon Aurora with MySQL-compatibility, a relational database built for the cloud. With RDS Performance Insights SQL-level metrics, you can identify high-frequency, long-running, and stuck SQL queries in seconds.  

Amazon Elastic Compute Cloud (EC2) now supports the use of custom identifiers to reference Amazon Machine Images (AMIs) during instance launch. You can create these identifiers using AWS Systems Manager Parameter Store and use them to reference your AMI during instance launch, reducing likelihood of instance misconfiguration and simplifying maintenance of your automation. Code that references these parameters no longer needs to be modified every time a new version of an AMI is created.

You can use Amazon CloudWatch to monitor Prometheus metrics from Amazon Elastic Kubernetes Service (EKS) and Kubernetes clusters, now available in beta. With this new feature, DevOps teams can automatically discover services for containerized workloads such as AWS App Mesh, NGINX, and Java/JMX. They can also expose custom metrics on those services, and ingest the Prometheus metrics in CloudWatch. By curating the collection and aggregation of Prometheus metrics, CloudWatch users can monitor, troubleshoot, and alarm on application performance degradation and failures faster while reducing the number of monitoring tools required.

We’re excited to announce recommendation reporting changes to Amazon CodeGuru Profiler that lets customers respond to issues quickly. With this new change, recommendation reports can now be run on an hourly basis instead of once daily. 

Amazon Pinpoint is now available in the AWS GovCloud (US-West) Region. With Amazon Pinpoint, you can engage your customers by sending marketing campaigns and transactional messages through email, SMS, push notification, and voice channels.