Evrone developed a high-load data processing system for SberMarketing, capable of handling 100,000 requests per second. Designed in Python with Granian for speed, the service standardizes, enriches, and stores event data before transferring it to long-term storage. Testing showed Python+Granian matched Go's performance, and a Tarantool-based Data Lake ensured data reliability. The client now has a scalable, maintainable solution.
Read the full case study here.
Top comments (0)