Amazon ElastiCache now allows you to assign permissions to specific resources in AWS IAM policies. You can now assign an IAM principal permissions to specific ElastiCache resource or resources. Prior to this release, Amazon ElastiCache did not support resource level permissions; customers could only assign permissions to all resources for a given action. With this release you can be fine-grained in your IAM policies and allow access to specific ElastiCache resources. For example, you can allow administrators in your organization to create production ElastiCache clusters and restrict other principals from modifying those specific clusters. This provides you with the flexibility to meet your enterprise security and compliance standards.

AWS License Manager now allows customers to enforce assignment rules when they bring their own licenses to EC2 Dedicated Hosts. Software licenses such as Windows Server and SQL Server often require customers to keep the licenses assigned to dedicated hardware for a specified period. Customers can now manage such assignment rules using AWS License Manager by specifying the period for which the license should remain assigned to EC2 Dedicated Hosts. 

Today, AWS Copilot CLI for Amazon Elastic Container Service launched version 0.3.0. Starting with this release, you can configure details about an AWS Copilot environment such as a pre-existing VPC, subnets, and CIDR ranges, allowing you to use infrastructure created outside of AWS Copilot. Additionally, you can now configure how AWS Copilot builds your services with overrides, such as arguments to use in a Dockerfile and the ability to set a context for a Dockerfile. Finally, you can now run one-time tasks or batch jobs with the new task run command; this is useful for containers that do not always need to be running or restarted such as running a script or querying a database.

Amazon Elastic Container Service (ECS) today launched the new Amazon ECS Optimized Inferentia Amazon Machine Image (AMI), a new Amazon Linux 2 based AMI for the Amazon EC2 Inf1 Instances on ECS. This makes it easy for customers to run Inferentia based containers on ECS using the ECS Optimized Inferentia AMI which comes pre-baked with all the necessary AWS Neuron packages. 

Amazon Textract is a machine learning service that makes it easy to extract text and data from virtually any document. One advantage of services like Textract is that customers benefit from continuous improvement over time. Today, we are pleased to announce one such quality enhancement to our table recognition feature. The new model detects rows and columns of large tables that span an entire page more accurately. Overall table detection and extraction of data/text within tables has also been improved.

Amazon Connect now returns agent to previous status (e.g. Available, Offline, Custom status) after hanging up an outbound call and completing after contact work from the Contact Control Panel (CCP). Previously, agent status was automatically set to Available even if agent had been in a Custom or Offline state before making the outbound call. 

AWS Site-to-Site VPN now supports IPv6 traffic for VPN connections to AWS Transit Gateway. With this launch, you can now use IPv6 addressing for traffic between your customer gateway device and your resources in AWS. If you use IPv6 addresses for your resources in Amazon Virtual Private Clouds and on-premises data center or office sites, you can now communicate securely between them using AWS Site-to-Site VPN connections.  

Starting today, you can enable the native desktop experience for your Amazon AppStream 2.0 fleets. Previously, AppStream 2.0 streaming sessions only showed the windows of the applications that were launched. With this launch, you can enable Desktop View on your fleet, which provides the same desktop experience your users are used to with the existing non-persistent capabilities AppStream 2.0 provides. You can enable the Desktop View when you create a new fleet, or update a stopped fleet. To get started, see Create a Fleet in the Amazon AppStream 2.0 Administration Guide . 

AWS Transfer Family now offers Federal Information Processing Standards (FIPS) 140-2 compliant endpoints in US and Canada commercial AWS Regions to protect sensitive information. These endpoints terminate Transport Layer Security (TLS) sessions using a FIPS 140-2 compliant cryptographic software module, making it easier for you to use Transfer Family for regulated workloads.  

Amazon Comprehend is a natural language processing (NLP) service that uses machine learning to analyze text documents and identify insights such as sentiment, entities, and topics from text. No ML expertise required. You can use Custom Entity Recognition to identify terms that are specific to your domain. For example, you can instantly extract product names, financial entities or any term relevant to you from unstructured text documents. Starting today, Amazon Comprehend is adding support for the following five new languages to Custom Entity Recognition: French, German, Italian, Portuguese, and Spanish.