Case Study: Powerspace achieves ultra-stable, low-latency analytics and ~25% infrastructure savings with Google Cloud Platform

Powerspace Creating an ultra-stable stack for a more productive team

Powerspace, a Paris-based adtech company that serves about 1 billion ads per month, needed to process several terabytes of data hourly with extremely low latency and near-constant availability to deliver real-time, personalized advertising in email newsletters. Its on‑premises data center was unreliable and time-consuming to maintain, causing occasional message loss, unexpected dropouts and a heavy operational burden on the engineering team.

Powerspace migrated to Google Cloud using Cloud Pub/Sub, Dataflow, BigQuery and Google Kubernetes Engine, creating a parallel data flow to move systems with zero downtime and no lost messages. The move delivered a rock‑solid, low‑latency stack (6 ms on Cloud Bigtable), freed developers from infrastructure maintenance, improved team productivity and confidence, and is projected to cut infrastructure costs by about 20–25%, enabling plans to expand and build SaaS and ML capabilities.

Powerspace

Cédric Sadai

Co-founder