A workload fork — one stressed warehouse becomes two specialized systems.
01 — The problem
An analytical warehouse was also serving high-frequency application lookups; continuous queries prevented effective scale-down and raised cost.
02 — The system
Select a layer to see its role.
Data flow
Technology — why it was used
AWS
Redshift Serverless
S3 Tables
Aurora PostgreSQL
DynamoDB
Athena / Presto
05 — Evidence
Application serving
Analytics workload
Takeaway
Match the engine to the access pattern — not every workload belongs in the warehouse.
What landed