Back to Portfolio
Project · Cloud Migration · 12 months

Migrating a 200-server data analytics platform to AWS under a fixed-budget, fixed-term mandate.

Project Management framework for a global financial services enterprise's pilot cloud migration — designed to prove AWS feasibility for a much larger enterprise modernization program. Twelve months. $1.2M fixed budget. Zero tolerance for unscheduled downtime.

RoleProject Manager
Duration12 months
BudgetUSD 1.2M
Scope~200 servers
Project ManagementCloud MigrationAWSAgile · 2-week sprintsRisk ManagementStakeholder GovernancePMI / PMBOK
01 — Context & Diagnosis

An aging footprint, rising risk, and a feasibility window that closes quickly.

The starting position

The client operated a complex legacy infrastructure built through decades of growth and acquisitions — spread across six aging data centers, with roughly 200 servers supporting the internally-facing data analytics platform. Operational overhead was high, technological risk was rising, and the cost base was no longer defensible.

Leadership had committed to a multi-year enterprise modernization. But before signing off on that broader program, they needed proof: a contained migration that demonstrates AWS feasibility, validates the operating model, and produces reusable patterns for the larger effort to follow.

The mandate

Move fast, but don't break the business. The data analytics platform had to keep running during core working hours — every minute of unscheduled downtime translated directly to financial loss and disrupted decision-making across the company.

The project also had to do double duty: deliver the migration and produce the SOPs, architecture decision records, and validated patterns the enterprise would later use to scale the same approach across the remaining infrastructure.

02 — Strategic Approach

A lift-and-shift, sequenced as ten iterative cycles.

Rather than treating the platform as one monolithic migration, I structured delivery around ten prioritized use cases — each migrated end-to-end inside a 2-week sprint cadence. Risk is localized. Value is demonstrated continuously. The team learns and tightens estimates with every cycle.

Infochart 01 · Delivery Methodology
Five-step migration flow, repeated per use case
Each cycle follows the same disciplined sequence, governed by a formal SOP and gated by UAT before any production cutover.
Use Case Migration Cycle · 2-Week Sprint01DiscoverInitial MappingDependency analysis,scope confirmation02MoveDatabase MigrationTransition data sourcesto AWS target03RebuildAnalytics MigrationProcessing & applicationlogic to new environment04ValidateWeb Server + UATEnd-to-end experience,mandatory UAT gate05CutoverDecommissioningFormal cutover, resourceretirementMandatory Quality Gate
03 — Master Schedule

The 12-month Gantt: phased setup, parallel use-case streams, hypercare close-out.

Use Case 1 functions as the pathfinder — extended timeline to absorb learning. Subsequent use cases run in parallel pairs to double throughput without doubling cost.

Infochart 02 · Master Schedule
12-month Gantt — phases, milestones, parallel streams
JanFebMarAprMayJunJulAugSepOctNovDecQ1Q2Q3Q4Planning & SetupInventory · Env. · PoCUse Case MigrationUC1 · PathfinderUC2UC3UC4UC5UC6UC7UC8UC9UC10Closure & HypercareHypercareM1Kickoff +Env. ReadyM2PathfinderCompleteM39 of 10UCs LiveM4FinalCutoverLegendSetup & foundational phasesUse-case migration sprintHypercare & closureExecutive milestone
04 — Budget & Resourcing

A fixed-rate, fixed-FTE envelope that absorbs scope variance without absorbing margin.

The financial model is deliberately simple: 800 hours per month at a blended rate of $125/hour, totaling exactly $100,000 in monthly burn. Predictable for the client, defensible internally, transparent under audit.

$1.2M
Total / 12 months
Fixed-fee professional services envelope
$6K
Per server migrated
Within the industry benchmark of $3K–$8K
9,600
Total billable hours
800 hours / month × 12 months
Infochart 03 · Resource Mix
Monthly resource allocation by role (800h envelope)
Engineering capacity dominates; management overhead held to ~12% to maximise delivery throughput.
Hours per month per roleCloud Engineer336h · 42%Data Engineer336h · 42%Cloud Engineering Tech Lead64h · 8%Project Manager64h · 8%Total monthly capacity800h = USD 100,000 / month @ $125/hr blended rateThroughput note2 parallel engineering streamsdouble output without doublingcost — 84% of capacity isdelivery-side, not overhead.
05 — Risk Management

Four risk domains, each with concrete mitigations and a measurable success state.

Infochart 04 · Risk Register
Risk · Mitigation · Target outcome
DomainRiskMitigationTarget OutcomeABusinessContinuityUnscheduled downtimeduring working hours• Pre-scheduled windows, 2-week notice• One use case at a time isolates impact• Mandatory UAT + rapid rollback ready• Sources stay live until validationZero unscheduleddowntime in hoursBProjectDeliverySpeed vs. qualitytension• Non-negotiable quality gates pre-prod• Fixed 2-week sprints, velocity tracked• Embedded technical writer for docs• UC1 as pathfinder, extended discovery100% UAT pass rate;zero critical incidentsCTechnicalDiscoveryLegacy systemcomplexity• PoC exposes hidden dependencies• Initial Mapping per use case• Real-time sharing in daily standups• AWS Discovery + traffic analysisComplete dependencymap; refined estimatesDBudget &TimelineCost overruns,schedule delays• Fixed-rate model ($125/hr) = predictability• 800h monthly burn, managed by velocity• Parallel streams 2× throughput, flat cost• Continuous monitoring in weekly PM reviewDelivered within$1.2M / 12 monthsRisk severityHighMediumLow— severity reflects pre-mitigation exposure.
06 — Communication & Team

A two-tier governance model that protects delivery focus and routes escalation cleanly.

Operational decisions stay with the working team. Strategic decisions — scope, budget, escalated risk — route to a defined steering committee on a defined cadence. The structure is deliberately small and the responsibilities are deliberately clear.

Infochart 05 · Governance Structure
Two communication tiers, one shared accountability
STRATEGIC TIER · Steering committee · Bi-weeklyDelivery OrgExecutive Lead+ ManagementProject ManagerScope · Time · CostRisk · CommunicationClient SponsorsHead of Data+ Tech LeadershipOPERATIONAL TIER · Working teams · Daily standups + bi-weekly sprint ceremoniesDelivery EngineeringCloud Engineering Tech LeadData Engineering Tech Lead2× Cloud Engineers2× Data EngineersJoint WorkgroupDaily standups (15 min)Sprint planning · ReviewsRetrospectivesSprint reports → PM → exec.Client TechnicalClient Tech LeadUAT validatorsSubject matter expertsSecurity & IT liaisonsEscalation pathscope · budget · risk
07 — Deliverables

What the client receives — and what the enterprise keeps.

The mandate has two layers: deliver the migration and produce the reusable artifacts that enable the broader enterprise rollout. Every output below was designed to outlive this engagement.

Project Charter & Scope Statement
Formal mandate, objectives, fixed-term and fixed-fee constraints, and explicit out-of-scope list.
Master Gantt Schedule
12-month timeline with phases, parallel streams, milestones, and dependencies tracked weekly.
Budget & Resource Plan
Fixed-rate envelope with monthly burn, FTE allocation, and per-server unit economics.
Risk Register & Mitigation Plan
Four prioritized risk domains with mitigations, owners, and measurable target outcomes.
Communication Plan
Two-tier governance, defined ceremonies and channels, sprint report template, escalation paths.
Team Structure & RACI
Roles, responsibilities, and decision rights mapped across both organisations.
Standard Operating Procedures
Reusable SOPs covering the five-step migration cycle, applicable to the enterprise rollout.
Architecture Decision Records
Pattern library and ADRs that constitute the operational blueprint for enterprise scale-out.

The project is the migration. The deliverable is the blueprint. A well-run pilot is the most expensive thing an enterprise can skip — and the cheapest one it can run.