Improving MiriCanvas's performance by 50% and costs by 30% with Amazon DocumentDB

MiriCanvas moved from MongoDB Atlas to Amazon DocumentDB to remove IOPS limits and reduce database costs.

After compatibility checks and query/index tuning, the team reported about 50% better performance.

The migration also cut costs by roughly 30% while supporting a real-time collaborative editing workload.

The case highlights how MongoDB-compatible services can ease bottlenecks in large-scale, concurrent editing systems.