Application Refactoring · Marketing Personalization
Meesho's rapid growth and peak-sale traffic overwhelmed its self-managed Kafka infrastructure, creating downtime risk, limiting throughput, and preventing the company from using customer and purchase data in real time for personalization.
Reported impact
Meesho can now handle much larger sale-day traffic, deploy new services faster, lower infrastructure overhead, and personalize shopping experiences with a real-time recommendation engine backed by current customer data.
Meesho adopted Confluent Cloud as the core data-ingestion layer for its e-commerce ecosystem so it could re-architect around elastic real-time streams instead of manually scaled Kafka clusters. The managed platform now absorbs bursty traffic during major sale events, carries data from across the marketplace, and makes those streams available to downstream services without the operational drag of cluster maintenance and over-provisioning. Because customer and purchase events are available immediately, teams can use the streaming layer for new retail experiences such as a real-time recommendation engine rather than waiting for delayed historical datasets. Confluent's pay-as-you-go model also lets Meesho size for actual bandwidth and usage instead of peak guesses. Together, those capabilities help the company ship features faster, survive traffic spikes, and turn fresh marketplace data into more relevant offers and product suggestions for shoppers.
Marketplace, customer, and purchase events stream into Confluent Cloud as the central ingestion layer, where downstream services consume them for peak-sale operations and for real-time recommendation and offer-generation use cases.
"Confluent Cloud took up position as the central ingestion layer, transforming the way data flowed through Meesho's ecosystem."
"It allowed the company to efficiently handle massive throughput bursts during peak events, such as the Diwali sale, where it reached up to 1 million requests per second."
"Now, with Confluent, Meesho has unfettered data access-which has helped the online platform create a real-time recommendation engine."
"Teams can handle 3x more load during sale days."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Alibaba Cloud Sample Archived · Oct 24, 2025 Drunk Elephant In 2019, Shiseido acquired Drunk Elephant , one of the fastest growing skincare brands in the last decade. Drunk Elephant is a highly effective skincare brand with research proven results, established in the United… Container Service for Kubernetes（ACK） · Model Studio · Tongyi Qianwen (Qwen) 
Confluent Sample Archived · Dec 17, 2025 Getir Getir needed to replace a distributed monolith, synchronous service calls, and stale ETL-fed machine learning data so rapid-commerce teams could scale checkout, discovery, and logistics without cascade failures. 40 million people Apache Flink · Confluent Cloud Confluent Sample Archived · Apr 24, 2025 Swiggy Swiggy needed to replace self-managed Kafka so it could scale delivery operations, predictions, and recommendations during peak demand without spending engineering time on infrastructure management. 680 cities Apache Flink · Confluent 