Phase 07 / 08 TRANSFORM

Deploy

From quality evidence to controlled production operations

The Deploy phase takes the tested solution into production with minimal risk: declarative infrastructure, GitOps-based delivery and progressive rollout strategies (canary, blue-green) with automatic rollbacks. Legacy replacements follow a staged cutover plan with parallel runs.

Position of the phase in the TRANSFORM delivery model

Process model

The phase delivers the productive target solution with proven operability: an automated release track, working rollback paths, a completed cutover and a documented system handed over to operations.

Release & cutover planning
Infrastructure & environments (IaC)
Progressive delivery
Data migration & parallel run
Cutover & handover to operations
1

Release & cutover planning

Define rollout strategy, cutover windows, data migration steps and abort criteria per application; align with business, operations and change management.

2

Infrastructure & environments (IaC)

Declarative provisioning of all environments as infrastructure as code; environment parity from test to production, secrets and certificate management.

3

Progressive delivery

GitOps-based deployments with canary or blue-green strategy; automated health checks and metric-based promotion or rollback.

4

Data migration & parallel run

Staged data migration with reconciliation; parallel run of legacy and new system until equivalence is proven at production scale.

5

Cutover & handover to operations

Controlled cutover with a hypercare phase; handover of runbooks, dashboards and escalation paths to the Operate phase.

Methodology mix

Low-risk go-live through progressive delivery

MethodPurpose in the TRANSFORM contextReference
Continuous deployment & deployment pipelineFully automated release track with staged quality gates; foundation of short, low-risk release cycles.Humble & Farley (2010)
GitOpsDeclarative desired state in version control with automatic reconciliation; auditable, reproducible deployments.OpenGitOps / CNCF
Progressive delivery (canary, blue-green)Stepwise traffic shifting with metric-based promotion; limits the blast radius of failures.Google SRE
Infrastructure as codeVersioned, testable infrastructure definitions; eliminates configuration drift between environments.Morris (2020)
Strangler fig migrationIncremental replacement of legacy systems behind a facade; avoids big-bang cutovers.Fowler (2004)
DORA change managementEmpirically validated practices for low-risk changes; measured via change failure rate and MTTR.DORA / Forsgren et al.

Platform support

ReqPOOL Suite: Signoff (AI-based acceptance) · REAM (Enterprise architecture)

Signoff — user acceptance & sign-off

Guided playbooks for business users with multimodal evaluation of the screen recordings; an independent AI verification model assesses every test case — the final acceptance stays with the client.

Signoff — acceptance report as go/no-go basis

The revision-proof acceptance report with full traceability from specification clause to test result is the documented basis of the go-live decision — robust towards suppliers and auditors alike.

REAM — cut-over steering

Cut-over plan, data migration and interface switching; parallel operation is reconciled against the AS-IS baseline.

REAM — controlled legacy decommissioning

Decommissioning of legacy systems on the target date; the landscape inventory is updated with the go-live.

Platform usage best practices

  • Start the cutover only once the Signoff acceptance report is available without open critical findings (gate: acceptance granted).
  • Define and technically rehearse abort criteria and rollback paths before every rollout — an untested rollback is not a rollback.
  • Reconcile the parallel run against the AS-IS baseline from the Analyse phase; unresolved deviations block the legacy decommissioning.
  • Document go/no-go decisions exclusively in the cutover protocol — verbal approvals are inadmissible.
  • Agree the hypercare period and staffing before the cutover; hand over to regular operations only with accepted runbooks.
  • After go-live, update the target landscape in REAM and follow up the legacy decommissioning to the target date.

Artifacts & outcomes

  • Automated release track with progressive rollouts
  • Infrastructure as code for all environments
  • Completed data migration with reconciliation evidence
  • Cutover protocol with go/no-go decisions
  • Operations handover package (runbooks, dashboards, escalation paths)

Quality gate — Deploy

Transition to the next phase happens through a formal quality gate (go/no-go). Gate criteria are documented in the gate review and signed off by the engagement lead.

  • Production stable over the agreed hypercare period
  • Rollback capability proven
  • Data reconciliation without open critical deviations
  • Operations handover confirmed by the Operate team

Scientific deep dives

Humble & Farley: Continuous Delivery (2010)

Standard reference for deployment pipelines and release management.

https://continuousdelivery.com/
Google SRE Workbook: Canarying Releases

Practice guide for metric-based progressive rollouts.

https://sre.google/workbook/canarying-releases/
OpenGitOps (CNCF)

Principles and standards of the GitOps approach.

https://opengitops.dev/
Fowler: Strangler Fig Application

Pattern of incremental legacy replacement.

https://martinfowler.com/bliki/StranglerFigApplication.html
Kim, Humble, Debois & Willis: The DevOps Handbook

Comprehensive reference on DevOps principles and flow optimization.

https://itrevolution.com/product/the-devops-handbook-second-edition/
DORA: capability catalogue

Research-based capabilities for delivery performance.

https://dora.dev/capabilities/

Best-practice guide: Deploy phase

Compact checklist covering process model, methodology mix and platform best practices for use in your engagement.

Download the best-practice guide (PDF)