Timeline view of data ingestion, audience, activation etc

Hi ​@Pradeep-Jaiswal 

I do not think there is one single AEP/AJO API that gives a complete E2E timeline across source ingestion → Data Distiller copy → Data Distiller transformation → audience evaluation → campaign execution.

The practical approach is to build a small operational timeline by stitching together the run metadata from each layer:

Source ingestion / AEP-XDM dataset
Use Flow Service monitoring APIs for source dataflow runs. For example, you can retrieve runs by flowId using the /runs?property=flowId=={FLOW_ID} pattern and capture start time, end time, status, record counts, and errors.

Batch / dataset confirmation
Use batch/dataset metadata and, where useful, ingestion events. Adobe ingestion events can show milestones such as data lake load, identity graph load, and profile service load, which helps distinguish “landed in dataset” from “available in Profile.”

Data Distiller copy and transformation jobs
Data Distiller jobs are scheduled Query Service jobs, so use Query Service schedules and scheduled-query-runs APIs. The useful endpoints are /schedules and /schedules/{SCHEDULE_ID}/runs. These expose run status and timing for each scheduled query.

Audience evaluation
Use the Segmentation Service API. /config/schedules gives the batch segmentation schedule, and /segment/jobs gives actual segment job execution status, timing, audience size, and errors. Batch segmentation schedules are generally once per day, so this is important when validating freshness.

So the recommended design is to create an internal “pipeline control table” or dashboard with the above APIs/jobs and sort by timestamp. That will give you the E2E operational timeline.

Also, I would not rely only on scheduled times. The real issue usually comes from actual completion time and dependency lag. For example, the audience job may be scheduled at 8 AM, but if the Data Distiller transformation completes at 8:10 AM, the audience evaluation may run on stale data unless the schedules are sequenced with enough buffer or triggered after the upstream job completes.