Case Study: PhonePe scales its data platform by 2000% and saves $5 million with Acceldata
Key results
The challenge
PhonePe, a Walmart subsidiary serving more than 350 million consumers across India and processing roughly $400 million in monthly transactions, faced scaling and performance issues on its open-source Hadoop environments. Rapid customer acquisition required quickly expanding the Hadoop footprint while adding Hive LLAP, Spark 3.x, and Druid, and existing tools could not support the growth and critical data initiatives.
The solution
Acceldata's data observability platform began providing real-time insights into HBase, Hive, Kafka, and Spark within 24 hours of deployment. The PhonePe team immediately identified problems with HBase region servers and tables under pressure, and Acceldata helped distinguish cluster issues caused by hardware or poorly designed tables from anomalies driven by seasonal and campaign-related surges, using multi-layer observability across infrastructure, platform, processing, and data layers.
“Acceldata supports our hyper-growth and helps us manage one of the world's largest instant payment systems. PhonePe's biggest-ever data infrastructure would never have been possible without Acceldata.”
BEBurzin EngineerFounder & Chief Reliability Officer, PhonePe
The results, in context
PhonePe scaled its data infrastructure rapidly from 70 to more than 1,500 Hadoop nodes, more than 2000% growth, and delivered 99.97% availability across its Hadoop infrastructure. It reduced data-warehouse costs by 65% by eliminating the need for commercial licenses and scaled its open-source data platform by 10x, and the case study reports total savings of $5 million.