My contribution
I shaped this portfolio exercise around a business problem, explicit acceptance rules and reviewable exceptions. The implementation and documentation were developed with AI assistance. It demonstrates an approach; it is not a record of a former employer's delivery.
The approach
The lab separates message validation, identity checks and delivery state. It uses a small, explicit subset of HL7 v2.5 ORM^O01 messages so the rules can be inspected and tested.
Decisions and tradeoffs
Treat a repeated message as the same event
Idempotency prevents a retry from creating a second order.
Keep identity conflicts visible
Conflicting patient or accession identifiers need a review path rather than a silent overwrite.
Make interruption recovery explicit
An offline queue and retry action let a reviewer inspect delivery state.
What to try
These scenarios are acceptance checks you can inspect in the demo and its tests.
| Scenario | Expected behavior |
|---|---|
| Valid sample order | Accepted with its patient and accession identifiers. |
| Resubmit the same message | Recognized as a duplicate without creating another order. |
| Conflicting identifiers | Flagged for review; no silent reassignment. |
| Queue while offline, then retry | Pending delivery can be retried after the interruption. |
What you can inspect
Interactive message exercise, explicit business rules, automated tests, an audit export and an executive summary export.
What this demonstrates
A working example with explicit rules and repeatable checks. No client outcome, cost saving or production performance claim is inferred from this exercise.
Moving toward a real implementation
For a real engagement: agree message profiles and identifier ownership, use approved synthetic test messages, define reconciliation and monitoring, and obtain clinical, security and operational sign-off.
Resumen en español
Este caso muestra cómo revisar órdenes de radiología, detectar duplicados y hacer visibles los conflictos de identidad. La demo usa datos ficticios y no se conecta a sistemas clínicos.