1. Notification System for Model Training Completion using AWS SES


    To create a notification system for model training completion using AWS Simple Email Service (SES), we will follow these steps:

    1. Set up an AWS SES Domain Identity: This is the domain that will be used to send the emails. It is a necessary step for SES to verify that you own the domain and to prevent spam.

    2. Create an AWS SES Configuration Set: Configuration sets are used to publish email sending events to Amazon SNS topics or other AWS services. We will use them to monitor the email sending for successful deliveries, bounces, or complaints.

    3. Set up an AWS SES Event Destination: This is tied to the Configuration Set and will direct the information about the email sending events to an Amazon SNS topic.

    4. Create an Amazon SNS Topic: This is the communication channel used to send messages or notifications to subscribers or other applications.

    5. Subscribe to the SNS Topic: Subscriptions define which endpoints get notified when a message or notification is sent to the topic.

    Here's a Pulumi program written in Python that sets up the above resources. This program assumes you have AWS credentials configured for Pulumi and that your AWS account has SES and SNS permissions. If you need assistance with AWS credentials, please consult the AWS Setup Guide.

    import pulumi import pulumi_aws as aws # Replace the following with your domain name domain_name = "your-trainings-domain.com" # 1. Set up an AWS SES Domain Identity domain_identity = aws.ses.DomainIdentity("modelTrainingDomainIdentity", domain=domain_name) # Verify the domain identity (manual step - this typically requires adding DNS records to your domain's DNS settings) # The verification process is beyond the scope of this code and must be done through the AWS console or your domain's DNS settings. # 2. Create an AWS SES Configuration Set config_set = aws.ses.ConfigurationSet("modelTrainingConfigSet", name="ModelTrainingDeliveryStatus") # 3. Set up an AWS SES Event Destination event_destination = aws.ses.EventDestination("modelTrainingEventDestination", configuration_set_name=config_set.name, enabled=True, matching_types=["send", "delivery", "bounce", "complaint"], sns_destination=aws.ses.EventDestinationSnsDestinationArgs( topic_arn="arn:aws:sns:region:<account-id>:ModelTrainingCompletionTopic" # Replace <region>, <account-id> with your actual AWS region and account ID )) # 4. Create an Amazon SNS Topic sns_topic = aws.sns.Topic("modelTrainingCompletionTopic", name="ModelTrainingCompletion") # 5. Subscribe to the SNS Topic (replace the email address with your actual notification email address) email_subscription = aws.sns.TopicSubscription("modelTrainingCompletionEmailSubscription", topic=sns_topic.arn, protocol="email", endpoint="notify-me@example.com") # Replace with your email address # Export the ARN of the SNS topic to use it outside Pulumi if necessary pulumi.export("sns_topic_arn", sns_topic.arn)


    In this program:

    • We start by creating an SES Domain Identity, which will allow AWS SES to send emails using your domain.
    • Next, we set up a Configuration Set. This will allow us to define and monitor email sending events.
    • The Event Destination will be linked to the Configuration Set, and is used to inform an SNS topic about the sending events we are interested in monitoring (such as email deliveries or bounces).
    • We then create an SNS topic that will act as a centralized outlet for sending notifications.
    • Lastly, we subscribe to the SNS topic using an email endpoint. When a message is published to the topic, the endpoint will receive an email notification.

    Make sure to replace "your-trainings-domain.com" and "notify-me@example.com" with your actual domain used for sending the emails and your email address where you want to receive the notifications. The verification process for Domain Identity is manual, as it involves adding specific DNS records to your domain's DNS settings.

    Next Steps:

    After deploying this code with Pulumi, your notification system will be set up. Once your model training job is complete, you can send an email from your domain using SES to notify you of the completion. If the email address in the subscription is verified, SES will deliver notifications to it.

    Keep in mind that AWS has a sandbox environment for SES that has certain limitations, such as the need to verify the recipient email addresses. To move out of the sandbox and use SES at scale, you need to request a sending limit increase and move out of the SES sandbox via the AWS support center.

    Remember to handle the bounce and complaint notifications properly to maintain a good sender reputation.

    For more information on AWS SES and how to manage domain verification, sending emails, and handling bounces and complaints, please refer to the AWS SES Documentation.