Case Study: AppsFlyer achieves 75% lower compute costs with Upsolver

AppsFlyer Cutting compute costs 75% by moving from a data warehouse to Upsolver + Amazon Athena

AppsFlyer, a mobile ad attribution and marketing analytics platform, needed a better way to handle petabyte-scale support data while reducing the high cost and complexity of its in-house ETL and BigQuery-based analytics. The company’s homegrown Clojure ETL pipeline required significant engineering effort to manage, tune, and maintain, creating delays and a growing backlog in analytics delivery.

To solve this, AppsFlyer implemented Upsolver with Amazon S3 and Athena to automate data lake ingestion, preprocessing, transformation, and optimization. With Upsolver, AppsFlyer reduced compute costs by 75%, processed 120 billion events per day, and saved hundreds of engineering hours, while also speeding up time to value and making ad hoc analytics easier for support and data teams.

AppsFlyer

Avner Livne

Real-Time Application (RTA) Groups Lead