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
Define rollout strategy, cutover windows, data migration steps and abort criteria per application; align with business, operations and change management.
Infrastructure & environments (IaC)
Declarative provisioning of all environments as infrastructure as code; environment parity from test to production, secrets and certificate management.
Progressive delivery
GitOps-based deployments with canary or blue-green strategy; automated health checks and metric-based promotion or rollback.
Data migration & parallel run
Staged data migration with reconciliation; parallel run of legacy and new system until equivalence is proven at production scale.
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
| Method | Purpose in the TRANSFORM context | Reference |
|---|---|---|
| Continuous deployment & deployment pipeline | Fully automated release track with staged quality gates; foundation of short, low-risk release cycles. | Humble & Farley (2010) |
| GitOps | Declarative 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 code | Versioned, testable infrastructure definitions; eliminates configuration drift between environments. | Morris (2020) |
| Strangler fig migration | Incremental replacement of legacy systems behind a facade; avoids big-bang cutovers. | Fowler (2004) |
| DORA change management | Empirically 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
Standard reference for deployment pipelines and release management.
Practice guide for metric-based progressive rollouts.
Pattern of incremental legacy replacement.
Comprehensive reference on DevOps principles and flow optimization.
Research-based capabilities for delivery performance.
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)