This Guidance uses API Gateway integrated with Amazon CloudWatch to provide monitoring capabilities so that you can track API calls and latency. This Guidance integrates with AWS CloudTrail for API request logging. These features improve observability of API operations and their usage. You can also set up CloudWatchalarms to notify you of errors or threshold breaches in near real-time.

Read the Operational Excellence whitepaper

This Guidance lets you use AWS Identity and Access Management (IAM) to strengthen data security and access control through granular permissions. By using IAM to grant least-privilege permissions, you can make sure that users have access only to the specific resources they need.

This Guidance uses Amazon Pinpoint, which provides built-in scalability and redundancy so that you can reliably deliver SMS, email, and push notification messages to users. Amazon Pinpoint is a fully managed service that delivers reliable communication with users even during peak times.

This Guidance uses DynamoDB, which provides fast, consistent, single-digit millisecond latency. It is also a fully managed NoSQL database that provides automatic scaling (through its on-demand mode) and replication across Availability Zones, facilitating high-performance efficiency and low latency. This makes it ideal for a high volume of users trying to access the website and application.

This Guidance uses Amazon Personalize, which charges based on the number of requests to your endpoints, so you only pay for recommendations that are generated. It also automatically scales resources so you don’t over-provision capacity. Additionally, as a fully managed service, Amazon Personalize removes the need for you to manage machine learning infrastructure. It handles capacity planning, model training, and hosting, so you don’t have to provision resources up front.

This Guidance uses Lambda, which lets you run code without provisioning or managing servers. Its automatic scalability and its reuse of implementation environments optimizes resource usage. This minimizes energy usage because Lambda runs only the necessary compute for your workloads. Additionally, you save more energy if you use Lambda functions powered by AWS Graviton.