I think the biggest production risk is the data path complexity relative to the app’s size: Azure Front Door -> Static Web Apps -> API Management -> Azure Function -> SQL Server. I see why you chose it for security and future scale, but for a cashflow app at this cost point it adds multiple failure domains, latency, and operational overhead without clear resilience benefits. APIM and Front Door are powerful, but if they are not justified by multi-region routing, partner APIs, strict throttling, or zero-trust edge requirements, they can become expensive choke points. I’m also concerned that SQL Server is the least serverless part of the design. It can dominate cost and availability planning, and I don’t see mention of failover groups, backup retention strategy, private endpoints, or how the Function handles connection pooling and transient faults. If this app depends on external housing data, I’d also want a clear ingestion/caching pattern in Storage to avoid hammering the database and APIs. I would simplify the request path or explicitly design for HA, private networking, and database resilience.