Nothing technical is holding this repo on the three-week release. It either already deploys without downtime (AWS) or is a modern, well-tested service the automated stealth path can carry today. Already deploys as a container to AWS, which releases without downtime today. Some artefacts still deploy on-prem; they follow the on-prem path. Separating it from the three-week release is a process decision, not a technical one.

• 1 system(s) depend on this repo; during a rolling deploy they will talk to old and new versions at once, so responses must stay compatible
• Consumes WCF contract packages: never take a breaking client bump mid-feature, and treat obsolete warnings as migration debt

1. Adopt the four-stage discipline from the ZDD engineering standard: pre-deployment compatibility, deployment, feature activation behind a flag, post-release cleanup.
2. Make every change dual-version safe: backwards-compatible APIs (new major version for breaking wire changes), expand and contract for any schema change, and consumers tolerant of unknown fields.
3. Audit serialized data: anything in Redis, queues or shared caches must deserialize cleanly across adjacent versions.
4. Wrap behaviour changes in feature flags so code deploys dark and release becomes a business decision, not a deployment event.
5. Route deployments through the Workflow Engine's automated drain, deploy, health-check, return cycle (the path proven by the first automated production run in mid-2026).
6. Verify with area-scoped regression rather than the whole-platform cycle, and watch error rates after rollout with automated rollback on alert.

sqlserver:CustomerPortal is shared, so this system cannot change engine on its own: every sharer moves in convoy. Either agree the move with all sharers as one programme, or first split this system's data out behind an owned API, then follow the steps below.

• sqlserver:CustomerPortal is shared with 2 other system(s)

1. Inventory the schema and query surface of sqlserver:CustomerPortal: tables, stored procedures, triggers, cross-database joins and anything else engine-specific.
2. Confirm the target: PostgreSQL (relational model; keep transactions and joins, swap the engine).
3. Data access is not through an ORM, so port the SQL by hand: translate T-SQL constructs, replace stored procedures with application code or target-native equivalents, and add integration tests around the queries that matter first.
4. Stand up the target alongside SQL Server, backfill with an ETL pass (pgloader or an export/import job), then dual-write or replicate until parity is proven.
5. Move reads to the target, watch correctness and latency, then cut writes over and keep SQL Server read-only until confidence is earned.
6. Decommission the SQL Server database and remove the connection strings and firewall paths that pointed at it.

Part of this repo already builds containers for AWS while other artefacts still deploy to on-prem machines. Finish the job by moving the remaining artefacts (see the hosting evidence on this page) through the same pipeline template.

1. Identify the artefacts still deploying through deployment-group release pipelines (listed in the hosting evidence).
2. Move each to the shared ci.yml pipeline template with the container parameters, as the already-containerised artefacts in this repo do.
3. Disarm the old release pipelines once traffic is confirmed on the new deployment.