Routing two Azure regions is configuration. Deciding what happens to two writes to one record is design. Four write strategies compared on one simulated outage, what Cosmos DB does with a conflict, idempotent retries across regions, readiness probes for Front Door, and the global tier in Terraform.
Read more →Tag: Azure
Azure Container Apps or AKS for a Regulated Workload: Who Patches What
For a regulated workload the container platform decision is mostly about who patches what and who can prove it. A ledger of thirteen operational duties across Container Apps, AKS Automatic and AKS Standard, the same security baseline written twice in Terraform with tests, and a decision rule short enough to argue with.
Read more →One Contract, Two Clouds: A Network Baseline on AWS and Azure with Terraform
Do not write one Terraform module that hides the cloud. Share the contract instead: the same inputs, outputs, naming and contract tests over an AWS implementation and an Azure implementation, tested on Terraform and OpenTofu with mocked providers, with the cost of the switch-module design measured.
Read more →Azure OpenAI Structured Outputs for Clinical Coding: Free Text to FHIR Condition and Observation, with a Human Review Step
A model proposes coded findings from a clinical note inside a strict JSON Schema whose enums are an approved code subset; code checks every proposal against the note; a clinician accepts or rejects; and only accepted items become FHIR Condition, Observation and Provenance resources. C# on .NET 10 and Python.
Read more →FHIR Validation in CI: Profiles, Terminology and IPS Checks That Fail the Build
A build gate for FHIR conformance: the HL7 validator behind a Python wrapper and the Firely .NET SDK check resources against the IPS profiles in GitHub Actions, with deliberately broken test data that proves the gate refuses what it should, and a clear account of what it cannot see without a terminology server.
Read more →FHIR Subscriptions and Events on Azure: Keeping Downstream Systems in Step
Azure Health Data Services announces FHIR changes through Event Grid, at least once and in no guaranteed order. An Azure Function on .NET 10 that keeps a downstream projection correct through duplicates, reordering, corrections and deletions with one version-guarded write.
Read more →