Azure migration: move what matters, in the right order
The short answer first; the detail is below it.
Short answer: an Azure migration is not a lift-and-shift by default. It is a decision about what should keep running, what should change, and in what order — made before anything moves. I assess what you run today, choose the Azure target for each part, and sequence the move so the business keeps operating while it happens.
Most cloud projects don't fail because the technology is wrong. They fail because the scope was too big, the architecture was an afterthought, and nobody asked whether the thing needed to move at all. A migration is where that pattern is easy to repeat — the wrong workload, moved the wrong way, for a reason nobody can explain a year later.
Who it is for
Teams and founders running something that works, who have been told it has to move to Azure — or who can see the bill or the risk and want to decide for themselves. It suits non-technical founders and operators as well as engineering leads.
It is built for teams in healthcare and other regulated industries, where privacy, security and compliance constraints are part of the design brief rather than a hardening pass at the end. You don't need to be technical to start. You need to know what the system does for the business, who depends on it, and what cannot break.
What you get
- A target architecture. What each part of the estate should run on in Azure, and why — including the parts that should stay where they are.
- A realistic cost view. Not a licence quote in isolation: the migration effort, the running cost, and what it costs to own afterwards.
- A sequence. What moves first, what moves last, what can run in parallel, and how the business keeps working during the change.
- A move, rework or replace call. For each part: move it as it is, rework it, or replace it with something that already exists.
- Security and governance from the start. Zero Trust access and the Microsoft Cloud Adoption Framework as the baseline, not an afterthought.
- Build-to-run. CI/CD, environment strategy, observability and cost transparency, so what your team inherits is operable, not just deployed.
The Azure target is chosen per workload. App Service for the web tier, Azure SQL for relational data and Zero Trust access are common answers — but they are answers to specific constraints, not a template applied to everything.
The plan is written so you can hand it to any team — including the one you already have.
How it works
It starts with the Strategy Call: forty minutes, paid, where you describe the estate and the constraints. From there the work is an assessment and a plan, and only then a build — if a build is the right answer. The assessment also records what the current system costs to run, so the comparison is against a real baseline rather than a vendor's estimate.
- Assessment. What you run, what depends on what, where the data sits, and which constraints are fixed.
- Architecture. The Azure target for each part, the security and governance baseline, and an honest cost view.
- Plan. The sequence, the risks worth proving first, and the decision on what does not move.
- Build, if it earns its place. I architect and deliver on Azure, or lead the team that does. The risky parts are prototyped in code, pipelines and configuration before the design is committed.
What it is not
- Not a body shop. I work like an architect, not a pair of hands. If what you want is an hourly migration team, that is a different engagement.
- Not a compliance certificate. I design within Azure's compliance capabilities, alongside your compliance owners. I don't audit you, certify you or give legal advice.
- Not a mandate to move everything. Moving a system that should have been retired is not a migration win.
- Not a rewrite by default. Your estate is likely a mix of move, rework and replace, and that split is a design decision worth making on purpose.
- Not a big-bang cutover by default. The sequence exists so the business keeps running. If a single weekend cutover is genuinely right, that is a decision — not an assumption.
Next step
Bring the current setup, the constraints you cannot change, and any quotes you have been given. Forty minutes is enough to say whether the plan you have is the right one — and what to do first either way.