Work With Me

I move engineering teams off the hyperscaler onto infrastructure they own — phased, reversible, and costed before you commit. The cloud bill is where every engagement starts: we read it, find what's recoverable in place, and then decide — with numbers, not ideology — whether the deeper fix is optimising what you rent or owning it instead. I run my own Kubernetes-on-bare-metal cloud, so the costs of ownership are ones I pay myself.

What I do

Cloud bill rescue

Where the bill is actually going and how to bring it down without breaking anything — right-sizing, commitments, storage lifecycle, data transfer, architecture. This is the first, fastest engagement, and sometimes it's all you need. If that's true, I'll say so.

Cloud escape — the migration

When the maths says owning beats renting, I plan and run the move: phased (one workload at a time, cloud stays up until the owned side is proven), reversible (every phase has a way back), and costed before you commit (hardware, engineering time, and the new run-rate on paper first). Observability, Kubernetes, storage, networking — the whole stack, on machines you control.

Owned-infrastructure handover

The last phase of the move, not a standing arrangement: observability that watches itself, backups and DR that are tested rather than assumed, and cost governance that holds after I have gone. It ends with a handover — the run-books are yours, and your team operates the estate.

Two of these written up in full — including what broke. Pick the one that matches your doubt:

“Am I really overpaying, and can you prove it?” An $80,000/month observability bill at one terabyte a day, what replaced it, and what the replacement actually cost to operate. Read it →

“Won’t the migration break something?” I replaced the platform underneath my own running production, node by node, with rollback at every step. Zero data loss — and the failures that happened anyway. Read it →

How it works

The free 15-minute bill review. You share the bill, I tell you straight whether there's real recoverable spend and roughly how much. If you're already lean, I'll say so and point you on. No pitch.
A fixed scope. A defined piece of work with a clear deliverable — the bill rescue with a savings roadmap, a costed migration plan, or the migration itself, phase by phase.
The deliverable. The report and the numbers, or the change shipped and proven — whichever the scope calls for. Every migration phase ends with the before/after measured, not asserted.

Who it's for

Teams whose cloud bill grew faster than their traffic — and who suspect renting is the reason.
Companies at the scale where owning starts to beat renting: steady workloads, real egress, a bill that reads like a second payroll.
Teams without a dedicated platform or FinOps hire — where infrastructure is one engineer's third job.

Start with the free 15-minute bill review

Tell me what you're dealing with and I'll come back within a day with an honest read — real recoverable spend, or confirmation you're already lean. Prefer to grab a slot directly? Book the review.

Or just email me at [email protected].

Not sure yet? The blog is the long version of how I think about this — start with when owning beats renting or the observability teardown.