Starting today, PyTorch customers can use TorchServe , a new model serving framework for PyTorch, to deploy trained models at scale without having to write custom code.

With AWS Cost Categories (GA), you can now categorize your cost and usage information precisely to your organizational structure and cost allocation needs such as teams, cost centers, geography, applications, and more. Using Cost Categories, you can create these unique categories and then write rules on which costs belong to each group. After defining your unique category, you can view, track, and optimize costs using Cost Categories in the AWS Cost Management suite of products such as AWS Cost Explorer, AWS Budgets, and AWS Cost and Usage Report (‘CUR’).

Starting today, Amazon RDS for SQL Server supports SQL Server Analysis Services (SSAS) directly on RDS DB instances. You can run SSAS in the Tabular mode. There is no additional cost to install SSAS directly on your Amazon RDS DB instance. 

Today we announced the general availability of AWS Elemental Link, a device that connects a live video source, like a camera or other video production equipment, to AWS Elemental MediaLive for video encoding in the cloud. To learn more and to place an order, please visit https://aws.amazon.com/medialive/features/link/.

Amazon Redshift now supports multi-factor authentication (MFA). Customers can use MFA to provide additional security when authenticating to their Amazon Redshift cluster. 

AWS Elemental MediaPackage is a video origination and just-in-time packaging service that allows video distributors to securely and reliably deliver live streaming or on-demand content at scale. From a single video input, MediaPackage creates video streams formatted to play on connected TVs, mobile phones, computers, tablets, and game consoles. It makes it easy to implement popular video features commonly found on DVRs, such as start-over, pause, and rewind. The service can also protect your content using Digital Rights Management (DRM) technologies.

AWS OpsWorks for Puppet Enterprise now supports custom domains. Through the existing interfaces you can now provide a custom fully qualified domain name (FQDN), its matching SSL certificate, and SSL private key and have the OpsWorks server configured accordingly. After you have created a Canonical Name (CNAME) entry in your DNS management service to point to the internal OpsWorks endpoint, you will find your Puppet Enterprise Dashboard under a domain of your choice instead of the generic OpsWorks-created one. Also, this will now be the public endpoint of your OpsWorks server. By using a custom domain with an SSL certificate that you manage, you have an endpoint that stays the same regardless of the server being re-created: any node under management by the OpsWorks for Puppet Enterprise server will stay associated through this endpoint. 

Starting today, contact center admins can improve their Amazon Lex chatbots by configuring timeouts for each customer response in a conversation. The timeout determines how long the chatbot waits for the customer to finish speaking. For example, you might configure a short timeout for a yes/no question, but a longer timeout when asking for an address or credit card number, to give the customer more time to finish speaking.

We are pleased to announce the release of NICE DCV version 2020.0 with the following new features:

• Stylus and touch support on Linux server
• Surround sound 7.1 playback on Windows server to Windows native client
• Hardware acceleration and stylus support on Linux native client
• New API command to set display layout on server side
• Multi-monitor web client support on the new Microsoft Edge browser (version 79.0.309 or higher)

AWS Security Hub has released a new API action called BatchUpdateFindings, and we plan to deprecate the current UpdateFindings API. The UpdateFindings API only supported a few fields in the AWS Security Finding Format (ASFF), and wasn’t integrated with CloudWatch Events. The BatchUpdateFindings API fixes those issues and supports a much larger set of fields that can now be updated, such as severity, criticality, confidence, user defined fields, notes, and workflow status. Also, the fields that BatchUpdateFindings can update cannot be updated by finding providers. Those fields can only be updated by the customer or by SIEM/ticketing/SOAR tools that have access to this API action. This prevents finding providers from overwriting your updates. You can use the BatchUpdateFindings API to complete actions such as creating your own suppression rules, changing severity scores, and adding notes to findings. To learn more about this API, please visit our documentation .