Data Engineering for Partner Reporting, Privacy Compliance, and Ad Metrics

A major telecommunications carrier engaged Taller to modernize the data pipelines behind its advertising business, covering partner reporting, privacy compliance, and ad metrics.

The data team inside the client’s advertising business faced three demands at once: a high volume of partner reports to produce on tight schedules, privacy-compliance rules that varied from state to state, and ad-metric calculations that had to be both fast and accurate. The engineering challenge was producing all three at scale while keeping the data auditable, so any figure could be traced back and reproduced when a partner disputed it.

Taller’s data engineers automated the partner-report exports, updated the pipelines to handle state-level privacy compliance, and built workflows to track ad metrics and impressions at the volumes the business ran. Ad-tech data engineering at this scale ran on a familiar stack: Snowflake, Python, Airflow (a tool that schedules and coordinates the steps of a data pipeline), and AWS.

What separated a reliable pipeline from a fragile one came down to reproducibility, and it showed up in three places. Each partner report needed to be regenerable from a specific snapshot of the data warehouse, so a dispute could be settled against a defensible record. Each privacy filter had to apply the rules of the jurisdiction the customer was actually in when an event occurred. And each metric calculation had to handle a long tail of edge cases without collapsing into a mess of one-off rules. That operational discipline lived in how the Airflow workflows were structured.