Case Study: Convertale achieves scalable, cost-efficient real-time recommendations with Amazon Web Services

Convertale - Customer Case Study

Convertale is a Turkey-based startup that provides a real-time recommendation engine to help e-commerce sites personalize customer experiences and boost revenue. Originally hosted on dedicated on-premises servers, the company struggled to scale with client traffic, lacked the elastic capacity needed for large experiments, and wanted a cost-effective, open-source–friendly infrastructure without upfront capital expenses.

Convertale migrated to AWS, using EC2 with Elastic Load Balancing and Auto Scaling, Amazon Kinesis, DynamoDB, S3, EMR, RDS, Elastic Beanstalk and integrations like Jenkins and RabbitMQ. The move went live in two weeks and now processes over 500 million user behaviors per month, raised availability from 98% to 99.965%, eliminated deployment downtime and unexpected outages, enabled large on-demand EMR clusters for experiments, and cut operating costs by about 50%, allowing more competitive pricing.

Convertale

Ismail Arslan

Product Manager