Here’s the American English translation of your text:
—
The use of private work teams in Amazon SageMaker Ground Truth and Amazon Augmented AI (Amazon A2I) has transformed the way organizations generate datasets, ensuring quality and security. Thanks to the intuitive interface of the AWS Management Console, many companies are choosing to automate this infrastructure through Infrastructure as Code (IaC), enabling safer and more efficient deployments while reducing the likelihood of human error.
However, implementing a private work team via IaC faces several technical challenges, primarily due to the interdependencies of the services required for the initial configuration. To overcome these difficulties, a comprehensive solution has been created that allows for the programmatic creation of private work teams using the AWS Cloud Development Kit (AWS CDK). This solution includes the setup of an Amazon Cognito user pool, tailored to the specific requirements of each client, and relies on a GitHub repository that provides customizable examples to facilitate this integration.
The complexity arises from the mutual dependency between the private team and the Amazon Cognito user pool. The application of the user pool requires parameters that are only available after creating the private team, complicating the configuration process. Additionally, the domain name of the pool must remain constant, as changing it after the initial creation can lead to access issues.
To tackle these challenges, the solution employs various components of AWS CDK, including custom AWS CloudFormation resources, allowing for the orchestration of user pool creation in a coherent and managed manner. The architecture resulting from this solution consists of a series of resources that, although many are only necessary for the initial configuration, are crucial for ensuring that workers can access and complete their labeling tasks.
Once the setup is complete, team members can access the labeling portal, following an authentication process that includes email invitations, registration, and platform access. Workers receive clear instructions from the platform, facilitating their onboarding and operation.
To implement this solution, it is essential to follow a structured deployment procedure and fulfill the prerequisites, which include cloning the GitHub repository and appropriately utilizing the AWS resources and credentials. With these measures in place, once workers are invited and registered, they will be able to manage their labeling tasks efficiently.
The implementation of a private work team should be accompanied by best practices recommended for Amazon Cognito and AWS CDK, ensuring not only operational effectiveness but also security tailored to the specific needs of each organization, integrating additional AWS services as necessary.
—
If you need any modifications or further assistance, feel free to ask!
via: MiMub in Spanish