This page is the dedicated landing for our cloud refactoring programme. It runs alongside the main publication and brings together three things we keep adjacent on purpose: the latest briefings, expert analysis from our editor and consultants, and the four shapes our advisory work takes when readers ask us to bring it into their team.
Each links to the full briefing on the main archive. Pricing is informational; the briefings are written first and quoted afterward.
A field-tested teardown of how a mid-size SaaS team trimmed 38% of their EKS spend without freezing roadmap delivery.
How a four-person platform group at a logistics startup built developer experience without becoming a permanent gatekeeper.
A patient, year-long modernization of a Java 8 monolith for a regional retail platform, told without the usual triumphalism.
A careful, low-drama account of decommissioning a homegrown auth service and the eight months of preparation it required.
These are not headline takes. They are the working positions we bring into engagements and into editorial reviews.
Every quarter we are asked to translate a proposed refactor into an annualised cost-of-delay number. We have stopped doing it. The framing flatters the analysis but compresses the actual decision into a single, often misleading figure. We prefer a small set of qualitative pause points, each defended in writing.
Refactoring, rewrite, migration, lift-and-shift, replatforming — they each have specific scope and risk profiles. We name the work plainly. Where a team is using the wrong word for what it actually intends to do, we tend to spend the first session of the engagement re-naming the work, not designing it.
Per-service compute attribution and effective utilisation against committed-use discounts. That is the working set. Anything else gets opened only when these two raise a question we cannot resolve.
We do not custom-shape engagements beyond these four. The constraint exists for the same reason a publication has an editorial line: it keeps the work consistently good.
A focused, fixed-price look at one capability or platform surface, with a written recommendation and a defended set of next steps.
Best suited to: Pre-program scoping, post-incident retrospective, second-opinion review.
A capability-by-capability migration sequence drafted with the responsible engineers, including pause points and rollback notes.
Best suited to: Active modernization programs that need to defend pace internally.
A short, hands-on session with the platform group, run on your premises or remote, with a working artefact at the end.
Best suited to: Calibration of vocabulary, frameworks, or shared decision criteria.
A recurring monthly or bi-weekly cadence, with a named consultant and a refreshable scope document each quarter.
Best suited to: Engineering leaders who want continuous, light-touch external review.