Executive Summary
A company acquired a SaaS product from external developers and discovered they'd bought code - but not control. The infrastructure was a black box running on Azure Container Apps with no documentation, no access controls, and unknown costs.
We rebuilt the entire infrastructure from scratch on Kubernetes, migrated all workloads, implemented FinOps practices, and delivered complete documentation - giving the acquirer true ownership of what they purchased.
The Challenge
Post-Acquisition Reality
The acquisition closed, but the infrastructure remained a mystery. External developers still had access. Costs were unpredictable. Nobody knew how anything actually worked.
What They Inherited
- Zero documentation - No architecture diagrams, no runbooks, no deployment guides
- External dependencies - Original developers still needed for any changes
- Azure Container Apps limitations - Platform constraints preventing needed customizations
- Unknown costs - No visibility into what was driving cloud spend
- No deployment pipeline - Changes deployed manually with no audit trail
- Security concerns - Unclear who had access to what
Our Solution
Complete Rebuild
Rather than patch the existing mess, we rebuilt the infrastructure properly - designed, documented, and deployable from day one.
12-Week Transformation
Discovery & Assessment
Mapped existing infrastructure, identified all components, documented current state, assessed migration risks.
Architecture Design
Designed target Kubernetes architecture, planned migration strategy, defined CI/CD pipelines, established FinOps framework.
Infrastructure Build
Provisioned Kubernetes cluster, implemented networking, set up monitoring, deployed CI/CD pipelines, configured security controls.
Migration Execution
Migrated workloads incrementally, validated functionality, performed parallel running, executed final cutover.
Documentation & Handover
Completed all documentation, conducted training sessions, performed knowledge transfer, decommissioned old infrastructure.
Key Components Delivered
- Kubernetes cluster - Production-grade, fully managed, properly sized
- CI/CD pipelines - Automated deployments with approvals and rollback
- FinOps dashboard - Real-time cost visibility by service and environment
- Monitoring stack - Prometheus, Grafana, alerting configured
- Complete documentation - Architecture, runbooks, deployment guides
- Access controls - RBAC, audit logging, security hardening
Results
| Aspect | Before (Inherited) | After (Rebuilt) |
|---|---|---|
| Documentation | None | 100% documented |
| Deployment control | External dependency | Full internal ownership |
| Cost visibility | Unknown spend | Real-time FinOps dashboard |
| Platform flexibility | Container Apps limits | Full Kubernetes control |
| Time to production | Weeks (external) | Hours (self-service) |
| Security posture | Unknown access | RBAC, audit logging |
"We thought we bought a product. What we actually bought was a black box. Now we finally own what we paid for - and we understand every piece of it."
Key Takeaways
- Acquisition does not equal ownership - Code without infrastructure control isn't really yours
- Rebuild can be faster than fix - Sometimes starting fresh is more efficient than patching
- FinOps from day one - Cost visibility prevents surprises and enables optimization
- Documentation is infrastructure - Undocumented systems create external dependencies