@GidelA17565353 First, confirm the customer identifier mapping end to end: Tealium CIS to Pega pyCustomerID or your primary key, including case sensitivity and trimming, and test a few known CIS through the API or container call. Next, check the Tealium to Pega integration is actually calling the right container and channel context, and that requests include customer ID, direction, treatments, and timestamp; capture and review a live request and Pega response. Validate the action’s eligibility and engagement policies at the exact evaluation time, including applicability, start and end dates, arbitration, contact policies, and any segment or NBA constraints. Verify Interaction History writing: the data flow that persists results to IH is enabled, healthy, and not backlogged in Kafka or DDS, and that outcomes are recorded with the same customer ID, channel, and direction you query later. Run tracer and Decisioning logs for one CIS from request to IH write, and check pxInteractionHistory, IH Summary datasets, and report filters for time zone windows that might hide recent rows. Check batch jobs that summarize IH are running and not failing, and reindex OpenSearch if Insights or Reports show stale counts. For missing CIS, sync sources so both Tealium and Pega customer tables load from the same golden record and confirm profile stitching rules if multiple IDs exist. Finally, build a small reconciliation report: last 24 hours of Tealium sends versus Pega container responses versus IH writes by CIS, to pinpoint where drops happen and fix that exact stage.