Get Started

To connect your service to CUE, you need your app's authentication token and a user ID. Marketers define how to use the response and which conditions to apply. Developers call the Context API and implement the rules in your service.

Before you startCopy Anchor Link

Confirm with your Airbridge contact that CUE is enabled and prediction data is ready for your app.

Prepare your authentication tokenCopy Anchor Link

The Context API uses the app SDK token or web SDK token for your app to authenticate requests. In the Airbridge dashboard, go to [Settings]>[Tokens] to find the token. See Tokens for details.

Store the token in secure server or CRM settings.

Choose a user IDCopy Anchor Link

Send exactly one of the following three fields in a Context API request. Use an ID collected for the same app as the token. The value must be a non-empty string.

brazeDeviceId is different from Braze's external_id.

Connect your user IDCopy Anchor Link

The ID you use in the Context API must identify the user you want to look up. For SDK retrieval instructions and setup for Braze Device IDs and external user IDs, see User ID mapping in Check Event Tracking.

Prepare a test requestCopy Anchor Link

Choose a test user and device with user ID mapping set up, then open the app to prepare the ID for your request. Include exactly one ID field in the body described in Send your first request, and send it with the token for your app. Follow Check the response to review the result.

Define your use caseCopy Anchor Link

Define the screen or message to personalize, the conditions to apply, and the default behavior:

• Choose the screen or message to personalize. For example, you can show different products on the first screen to users with a high predicted likelihood of purchase.
• Choose the predictions to use in your personalization rules. Purchase and churn predictions describe different events, so choose the ones that fit your purpose.
• Define the default screen or message for users who do not meet a condition or whose API response is unavailable.

Send your first requestCopy Anchor Link

The request below looks up an external user ID. Replace abc-123 with an external user ID that your app sent to Airbridge. Send the request from your server or CRM.

curl --request POST --include \
  'https://api.airbridge.io/entry/inference/v1/context' \
  --header "Authorization: Bearer ${AIRBRIDGE_APP_TOKEN}" \
  --header 'Content-Type: application/json' \
  --data '{"airbridgeExternalUserId":"abc-123"}'

For other ID types, see Payload examples in the Context API reference. Include exactly one ID field in the request body.

Check the response and your service's behaviorCopy Anchor Link

After sending the first request, check that the API response is correct. Then test whether your service uses it as intended.

Check the responseCopy Anchor Link

A successful request returns HTTP 200. The example below shows a successful response.

HTTP/1.1 200 OK
Content-Type: application/json
X-Request-Id: 01915c8e-7a12-7d3f-8b2a-1f2e3d4c5b6a

{
  "requestId": "01915c8e-7a12-7d3f-8b2a-1f2e3d4c5b6a",
  "appId": 1234,
  "user": {
    "id": "abc-123",
    "type": "airbridge_external_user_id"
  },
  "context": {
    "lastTouchpoint": {
      "channel": "media_A",
      "campaign": "moving_service",
      "adGroup": "moving_quote",
      "adCreative": "moving_guide_video",
      "term": null
    }
  },
  "predictions": {
    "d3_purchase": {
      "probability": 0.723,
      "tier": "high"
    },
    "d3_churn": {
      "probability": 0.812,
      "tier": "high"
    }
  }
}

• Confirm an HTTP 200 response for a test user whose prediction data is ready.
• Confirm that appId identifies the target app and that user.id and user.type match the value and type of ID you sent.
• Confirm that your service reads the ad attribution fields it needs from context.lastTouchpoint and that the values match the attribution result expected in your test.
• Confirm that your service reads probability and tier from predictions.d3_purchase and predictions.d3_churn and applies the intended rules.

Check your service's behaviorCopy Anchor Link

• Confirm the intended screen or message appears both when a user meets a rule and when they do not.
• For rules that use installation attribution, confirm that your default rules apply when the context.lastTouchpoint object or a required field is null.
• Confirm that your service uses the default behavior after an API error, a timeout, or an unreadable response.

Use the default behavior when predictions are unavailable. Converting an error into probability 0 or tier low would confuse missing information with a low predicted likelihood. See Error Handling for how to handle each situation.

Get help with setupCopy Anchor Link

When contacting Airbridge, include the request time, requestId, HTTP status code, and error.code. If you did not receive a requestId, include the URL, method, and what happened.

Once the initial setup works, review your existing events in Check Event Tracking. Then explore Using Context to define specific personalization rules and default behavior.