Case Study: Kamino Retail handles 3x peak traffic at sub-15ms latency with Aiven for Apache Kafka and ClickHouse
Key results
The challenge
Kamino Retail, a French AdTech startup founded in 2023, relies on real-time data as the cornerstone of its retail media business, processing thousands of events per second and holding itself to a sub-50ms response-time SLA for ad requests. As it scales across Europe and toward the US, it must handle large traffic surges during peak events while keeping storage costs in check.
The solution
Kamino Retail built its platform on Aiven for Apache Kafka and Aiven for ClickHouse running on Google Cloud, using ClickHouse tiered storage to move data older than three months from SSD to lower-cost object storage.
“The process of receiving a request and sending an event to the Kafka cluster takes less than 15 ms in 99% of cases, despite the huge volume of data (>5 TB) we tracked in Clickhouse for our retail clients in our first 18 months.”
EVEmmanuel ValetteKamino Retail
The results, in context
Kamino Retail reported handling 3x traffic surges during peak events, with the process of receiving a request and sending an event to the Kafka cluster taking less than 15ms in 99% of cases, despite tracking more than 5TB of data in ClickHouse over its first 18 months. The company anticipates its object storage growing by a factor of 10 to 50 as the business expands.