Quickly Build End-to-End Integrations to SaaS Partner Event Sources and AWS Services with Amazon EventBridge

By Santiago Cardenas, Technical Lead, Sr. Partner Solutions Architect, AWS Serverless
By Emily Shea, Sr. Partner Development Manager, AWS Serverless

Amazon EventBridge makes it easy to connect applications together by delivering a stream of real-time data from custom sources, Amazon Web Services (AWS), and software-as-a-service (SaaS) applications.

The data can then be sent to a variety of targets like AWS Lambda, AWS Step Functions, Amazon Kinesis, and many more.

Amazon EventBridge also enables you to connect your applications with a range of SaaS partners without having to worry about building and maintaining custom infrastructure. Customers are using these capabilities to improve the scalability and reliability of their applications by building event-driven architectures rather than tight service coupling.

Some of the typical use cases we see today are:

• Auditing in near real-time, and archival of historical business operations events.
• Visualization and analysis of events for business intelligence and operational purposes.
• Automated alerting and remediation of application, service, and infrastructure systems.
• Connecting custom workflows with downstream consumers and legacy or on-premises applications.
• Supercharging applications with AWS artificial intelligence (AI) services to do things like media captioning and localization, identity verification, document indexing, personalization, and forecasting.

As more customers have started building end-to-end integrations with EventBridge, we are often asked about the best way to get started exploring the array of different possibilities for event sources and event types.

With features such as Schema Registry that include automated discovery and code bindings, we already make it easy to take a peek inside the event structure and format. However, many customers have asked for examples of end-to-end integrations for SaaS partners through different scenarios and use cases.

Partner Event Source Integrations

AWS introduced Amazon EventBridge partner event source integrations to showcase reference architectures and end-to-end Amazon EventBridge use cases. These can help you get started quickly with integrating SaaS partners in your own applications.

Fully open source, these solutions include code and AWS Serverless Application Model (SAM) templates that can be customized and extended to fit your application’s needs.

The reference architectures include AWS Quick Starts (with the SAM templates) and the code, on GitHub, to ease integration with SaaS partner event sources and AWS services.

Integrations with SaaS Partner Event Sources

The first release includes integrations with these AWS Partner Network (APN) Partners:

• Auth0
• Buildkite
• Shopify
• Zendesk

Auth0

Visualization Integration — Sends Auth0 events that match an Amazon EventBridge rule (for example, when an Auth0 user successfully signs in) to AWS Lambda, and stores event data in Amazon Simple Storage Service (Amazon S3). Event data is ready to be used as a data source for Amazon QuickSight or other visualization tools.

• See the Auth0 Quick Start
• Get the code on GitHub

Figure 1 – Auth0 integration.

Buildkite

Visualization Integration — Sends Buildkite events that match an Amazon EventBridge rule (for example, ‘Job Finished’) to AWS Lambda and stores event data in S3. An AWS Glue database table and Amazon Athena query are created to query your data and act as a data source for Amazon QuickSight or other visualization tool.

• See the Buildkite Quick Start
• Get the code on GitHub

Figure 2 – Buildkite visualization integration.

Pipeline Alerts — Sends Buildkite events that match an Amazon EventBridge rule (for example, ‘Build Finished’) either to Amazon Simple Notification Service (SNS) to send the pipeline alert as a text message or email, or to AWS Lambda, which calls a webhook endpoint, such as Amazon Chime or Slack..

• See the Buildkite Quick Start
• Get the code on GitHub

Figure 3 – Buildkite pipeline alerts integration.

Build Workflow — Buildkite events that match an Amazon EventBridge rule (for example, ‘Build Started’) trigger an AWS Step Functions state machine. The state machine first invokes AWS Lambda to save the event data to S3 for later analysis. It then invokes Lambda to publish a message to an SNS topic.

• See the Buildkite Quick Start
• Get the code on GitHub

Figure 4 – Buildkite build workflow integration.

Shopify

Image Processing — When a new product image is uploaded to Shopify, the event is passed through Amazon EventBridge to trigger an AWS Step Functions state machine. The state machine calls a series of AWS Lambda functions that use Amazon Rekognition to identify the image labels and create relevant image tags in Shopify (for example, identifying and tagging an image as a ‘hat’).

• See the Shopify Quick Start
• Get the code on GitHub

Figure 5 – Shopify image processing integration.

Zendesk

Sentiment Analysis — When a new ticket is created in Zendesk, the event is routed through Amazon EventBridge to trigger an AWS Step Functions state machine. The state machine calls a series of AWS Lambda functions that use Amazon Comprehend to analyze the ticket content sentiment and escalate the priority of negative tickets.

• See the Zendesk Quick Start
• Get the code on GitHub

Figure 6 – Zendesk sentiment analysis integration.

Ticket Translations — When a ticket comment is created in Zendesk, the event is routed through Amazon EventBridge to trigger an AWS Step Function state machine. The state machine calls a series of AWS Lambda functions that use Amazon Translate to translate the comment content and post a new comment on the ticket with the translation.

• See the Zendesk Quick Start
• Get the code on GitHub

Figure 7 – Zendesk ticket translations integration.

Attachment Processing — When an attachment is uploaded to a ticket in Zendesk, the event is routed through Amazon EventBridge to trigger an AWS Step Functions state machine. The state machine calls a series of AWS Lambda functions that use Amazon Textract and Amazon Rekognition to identify text and images in the attachment and post a new comment on the ticket with this information.

• See the Zendesk Quick Start
• Get the code on GitHub

Figure 8 – Zendesk attachment processing integration.

Integration with AWS Building Blocks and Patterns

The first release includes integrations with AWS Lambda, and Amazon Kinesis Data Firehouse and Amazon S3.

AWS Lambda — An Amazon EventBridge event bus to trigger a rule that evaluates all events and invokes an AWS Lambda function as a target for matched events. Can be configured for any SaaS partner event source.

• See the Lambda Quick Start
• Get the code on GitHub

Figure 9 – AWS Lambda integration.

Amazon Kinesis Data Firehose and Amazon S3 — Enables your Amazon EventBridge event bus to trigger a rule that evaluates all events and puts records on an Amazon Kinesis Data Firehose delivery stream. This acts as a target for matched events, and delivers them to an Amazon S3 bucket. Can be configured for any SaaS partner event source.

• See the Kinesis Data Firehouse and S3 Quick Start
• Get the code on GitHub

Figure 10 – Amazon EventBridge and Amazon Kinesis Data Firehose integration.

Conclusion

The preceding integrations consist of reference architectures that include AWS Quick Starts (including SAM templates) and code to help you quickly start integrating SaaS partners in your own applications.

We’re excited to hear from how you’re using Amazon EventBridge integrations. Please reach out to us at eventbridge-partners@amazon.com.