Integration steps

The Amplify Platform enables the full life cycle of Application Integration use cases.  The following are integration steps in a typical integration scenario.

1. Define the business problem

Often the business problem is framed at a high level. For example, "we want to optimize the lead conversion rate in our company." This evolves into a tangible action plan that articulates the need for integrating components such as a lead management system, a customer relationship management (CRM) system and a messaging system so sales people are notified when new users sign up and their contact data are available for follow-up.

With the outline of the problem defined and the involved entities identified, you can move to the next step.

2. Design - what information needs to be exchanged and when 

You must define what needs to be provided in which application. That is, what information must flow from which sources to which destinations. For example, when a new lead is added in Marketo, "I want to synchronize the name, address, phone number and email details to Salesforce, and I want a notification from Slack that contains the contact person's first and last names, email address and phone number, with a reference to the newly created sales entry."

3. Configure the application interfaces

With clarity on the information that needs to be exchanged, you can start building the integration pattern. This starts with configuring access to the involved applications. Following the example from the earlier steps, you end up with configured interfaces to the CRM, messaging and lead management systems. For integration, you can leverage one of the more than 170 prebuilt connectors or you can define your own.  You can use the prebuilt connectors in cloud-to-cloud, ground-to-cloud and cloud-to-ground integrations. You can use a configured connector in Integration Builder and API Builder. High level, the steps look like this:

  1. Configure your connector instances in Integration Builder or build your own custom connector. After completing this step, a proxy for the defined interface is generated in Amplify Central. See the Central documentation for details on using proxies.
  2. In Central publish your connector to the Catalog from the proxy registry. Once published, the interface is available for self-service. Users can subscribe to the published interface. This step is optional and only needed to use the connector in API Builder or a third-party API client.

In addition, you can use other methods for creating interfaces. You can leverage API Builder native connectors or use the Swagger plugin to generate a Swagger flow nodeThe generated Swagger flow nodes expose methods you can use in API Builder flows for communication with the API services described in your Swagger documents.

4. Build the integration process

Once interfaces to the source and target applications have been established, you can start building the integration flow for moving the data. Use Integration Builder for cloud-to-cloud integration patterns. You can use a combination of API Builder and Integration Builder for other integration types.


Your integration process starts in Integration Builder for cloud-to-cloud flows. The following are the steps.

  1. Create the flow template. The template contains the logical steps to execute in the flow.
  2. Create a flow instance from the flow template. In the flow instance you hook your connector instances to your flow.

Flows consist of a trigger that initiates the workflow and the actions that occur subsequently. Application Integration supports the following types of triggers.

Trigger When to use
Event When integrations must be triggered based an event occurring in the source application.
Connector request When an API call is made to a connector instance.
Request When integrations must be triggered through an API call  to an endpoint.
Scheduled When integrations must run at specific times.
Manual When integrations are triggered manually via an API call.

See Integration Builder documentation for more information about connectors and designing flows.

Ground-to-cloud, cloud-to-ground and ground-to-ground

The following is a typical path.

  1. Create your project, during which you import the predefined connectors from the catalog.
  2. Create your integration flow.
  3. Create your application and run your application locally. Or, build, deploy and run your integration flow in a Docker environment.

You must manually trigger flows created with API Builder.

See the API Builder documentation for more information.

5. Test, deploy and run

With the integration processes defined, you can validate the functionality by running the flow in combination with non-production endpoints and interfaces. Depending on the trigger type, enter the information in the source application that triggers the event. Or, if this is a flow that is manually triggered, call the API to trigger it. 

For Integration Builder flows, once the obtained results are satisfactory, you can move the integration flow from the build environment to the next stage. This can be done in the user interface by exporting from the source environment and importing in the target environment or via the API. Connectors and flow templates can be exported and imported between environments. However, you must define connector and flow instances manually in each environment or through the API.

Related Links