How to Get Started with Amazon SageMaker

Getting started with Amazon SageMaker may seem complex at first, but the process is actually straight forward once you understand the key steps. In general, you only need to set up your environment and create your first project before you can begin building machine learning and AI solutions. 

Prerequisites

Before setting up Amazon SageMaker, make sure you have an active AWS account and proper access configured. It is recommended to: 

• Create an administrative user instead of using the root account  

• Enable IAM Identity Center for secure access  

• Set up multi-factor authentication (MFA) for better security  

These steps ensure your environment is secure and ready for production use. 

Create an AWS Account

Before using Amazon SageMaker, you need an account in Amazon Web Services. 

If you don’t have an account yet: 

1. Open https://portal.aws.amazon.com/billing/signup 
2. Follow the online instructions. 

Part of the sign-up procedure involves receiving a phone call or text message and entering a verification code on the phone keypad. 

When you sign up for an AWS account, an AWS account root user is created. The root user has access to all AWS services and resources in the account. As a security best practice, assign administrative access to a user, and use only the root user to perform task that require root user access.

Create a user with administrative access

After you sign up for an AWS account, secure your AWS account root user, enable AWS IAM Identity Center, and create an administrative user so that you don’t use the root user for everyday tasks. 

Secure your AWS account root user  

1. Sign in to the AWS Management Console as the account owner by choosing Root user and entering your AWS account email address. On the next page, enter your password. For help signing in by using root user, see Signing in as the root user in the AWS Sign-In User Guide.  
2. Turn on multi-factor authentication (MFA) for your root user. For instructions, see Enable a virtual MFA device for your AWS account root user (console) in the IAM User Guide.  

Create a user with administrative access  

1. Enable IAM Identity Center by following the instructions in the AWS IAM Identity Center User Guide.
2. In IAM Identity Center, grant administrative access to a user. If you are using the default IAM Identity Center directory as your identity source, you can configure user access based on your organization’s needs.
3. Sign in using the IAM Identity Center user by accessing the sign-in URL sent to your email during the user creation process. You can then log in to the AWS access portal and start managing your environment.

Assign access to additional users  

1. In IAM Identity Center, create a permission set that follows the best practice of applying least-privilege access. For detailed steps, refer to the AWS IAM Identity Center User Guide.
2. Assign users to a group and grant single sign-on (SSO) access to that group. For more information, see the AWS IAM Identity Center User Guide on managing groups.

Setting up Amazon SageMaker

Step 1 – Create an Amazon SageMaker unified domain  

The following steps guide you through creating an Amazon SageMaker unified domain using the Quick setup option.

1. Navigate to the Amazon SageMaker console at https://console.aws.amazon.com/datazone and select the appropriate AWS Region using the region selector.
2. Choose Create a Unified Studio domain, then select Quick setup.

With this option, you are creating an Amazon SageMaker unified domain while allowing SageMaker to automatically configure default capabilities that can be customized later. These include data analytics, machine learning, SQL, and generative AI, as well as data and AI governance features. It also supports generative AI application development using Amazon Bedrock serverless models, provides access to Amazon Q (Free Tier), and enables authentication via AWS IAM or AWS IAM Identity Center.

3. If you see a notification that no VPC has been set up, you can choose to create a new VPC (recommended) or use an existing properly configured VPC.

If you plan to use your own VPC, SageMaker Unified Studio allows you to select VPCs within the same account or shared VPCs from other accounts within your AWS organization.

4. If no models are accessible, you can select Grant model access to enable access to Amazon Bedrock serverless models for use in SageMaker.
5. Expand the Quick setup settings section and review the configuration. You can keep the default settings, then choose Continue to proceed with domain creation.
6. On the IAM Identity Center page, create a new SSO user or select an existing one to access SageMaker Unified Studio. This user will act as the administrator, since IAM roles cannot be used to log in directly.
7. Choose Create domain.

After a short time, an email will be sent to the address associated with the IAM Identity Center user. This email will prompt you to set a password and access your domain.

Step 2 – Create a new project  

In Amazon SageMaker, projects enable a group of users to collaborate on various business use cases. Within projects, you can manage data assets in the Amazon SageMaker catalog, perform data analysis, organize workflows, develop machine learning models, build generative AI apps, and more.  

Navigate to Amazon SageMaker Unified Studio  

To begin creating a project, navigate to Amazon SageMaker Unified Studio. You can do this by using the link in your email that you used to set an IAM Identity Center password, or by selecting the domain in the Amazon SageMaker management console and choosing Open unified studio.  

Sign in using your SSO credentials that you configured using the email from IAM Identity Center. If your IAM Identity Center is configured to require multi-factor authentication (MFA), set up and use an MFA device. Follow the instructions on the screen to register or use an MFA device as needed, or contact your admin for support. For more information about configuring MFA device enforcement, see Configure MFA device enforcement in the IAM Identity Center User Guide. 

Project name and description  

After navigating to Amazon SageMaker Unified Studio, choose Create project.  

The project name and description includes the following fields: 

• Project name – the name of your project. Enter a name here. The name of the project can not be edited after the project is created.  
• Description – an optional description of your project. You can edit this later.  
• Project profile – project profiles define which resources and tools should be provisioned in the project. These include tools and compute resources for SQL, data science, data engineering, and machine learning development. Project profiles can include resources and tools from Amazon Redshift, Amazon SageMaker AI, and other AWS services. To complete the use cases in this getting started guide, choose the All capabilities project profile.  

Choose Continue to review parameters. 

Review parameters  

On the next page of project creation, you can review and optionally edit the names and values for different resources that are created when the project is created. You can leave all the defaults and then choose Continue.  

Review  

Use the last page of project creation to review the configurations you have selected. When everything is configured as desired on the project creation review page, choose Create project. You are then redirected to the project home page. The project will start building and a progress bar will appear with the status. 

Source: AWS Documentation