Akamai Integration

DataStream 2 pushes edge request logs to Airefs every 30 or 60 seconds, where bot traffic is classified and assets and error responses are dropped server-side. This powers AI Impressions; to also track Clicks, add the client-side tracking script from Site Settings → Tracking Script.

Your access token is in your Airefs account under Site Settings → Access Token.

Step 1: Create the stream

1. In Akamai Control Center, open Common Services → DataStream and click Create stream.
2. Name the stream, pick the group and contract, and select the properties to monitor.

3. On the Data sets tab, choose Log format: JSON and select at least: reqTimeSec, reqId, cliIP, statusCode, reqMethod, reqHost, reqPath, queryStr, UA, referer, rspContentType, country.

4. On the Delivery tab, choose Custom HTTPS with endpoint URL:

https://api.getairefs.com/v1/logs/akamai

5. Authentication type BASIC: username airefs (any value works), password = your access token. Akamai does not allow Authorization as a custom header, which is why the token travels as the Basic auth password.
6. Content type application/json. Compressed data is optional; both work.
7. Set the push frequency and click Validate & Save. Akamai sends a validation request; it succeeds when the token is correct.

Step 2: Activate the DataStream behavior

1. In Property Manager, edit your property and add the DataStream behavior, selecting your new stream.
2. Activate the property version on production. Logs flow once both the stream and the property version are live.

Optional but recommended: enable the Log Request Details behavior (populates referer) and include query parameters in the cache key (populates queryStr) for better attribution.

Verification

Visit a few pages on your property after activation. Events appear in your Airefs dashboard within a few minutes. To check parsing before going live, POST a sample log line to the endpoint with ?dry_run=1 — Airefs returns a per-record verdict including the detected bot classification, without storing anything.