We don't disappear after launch

The unglamorous work that prevents the expensive disasters: done on schedule, documented, and reported to you monthly.

• Updates little and often, never a crisis

  Software updates applied in small, tested batches on a regular schedule, instead of being saved up into an annual emergency. It's what keeps a future upgrade a routine task instead of a rewrite.

• Critical patches within an agreed window

  Serious vulnerabilities are assessed against your actual setup and patched within a promised timeframe, with out-of-hours handling for anything actively being exploited. You get a written record of what was patched and when.

• Scanning for weaknesses, constantly

  Continuous automated scanning of everything your systems are built from, plus detection of passwords accidentally left in code. Findings are judged by real risk to you, not forwarded as raw alarm noise.

• Old keys and ex-staff access removed

  A scheduled review of who can access what, removing dormant accounts and tightening over-broad permissions. Former staff and forgotten integrations keeping live access is one of the most common real-world ways companies get breached.

• Backups proven by restoring them

  Backups are verified by actually restoring them on a schedule and timing how long it takes, so your recovery time is an observed fact, not an assumption.

• Never fixed live on the real system

  Every change, even an urgent fix, goes through a safe copy first, gets reviewed, and can be undone in one step. Slower on the smallest fixes, dramatically safer on everything else.

• A rehearsed plan for bad days

  A written playbook covering severity levels, who gets called, how we communicate, and the legal notification deadlines where privacy law applies: rehearsed rather than filed, so the first time we use it isn't the first time we read it.

Everything in writing, everything recoverable, you're never dependent on any one person's memory, including ours.

• Promises with penalties (SLAs)

  Response and fix times by severity, coverage hours, escalation contacts, and service credits when we miss, in the contract, not on a marketing page. Plus a named engineer who knows your systems, not a rotating queue.

• Documentation that stays current

  Diagrams, guides, account registers, and recovery procedures kept up to date as part of the monthly work, not written once at handover. Our test: another competent team could take over from the documents alone.

• A plan for the worst-case scenario

  Agreed targets for how fast you're back up and how much data you could lose, an off-site backup strategy, and a rehearsed failover. We tell you honestly what each level of protection costs, so it's your commercial decision.

• Monthly reports a director can read

  Uptime, incidents and their causes, work completed, patches applied, spend, and what we recommend next: written for a director, not an engineer. No jargon, and no invoice arriving without an explanation of what it bought.

• Hours every month to make things better

  A defined block of improvement hours each month (small features, refinements, polish), prioritised by you, rolling over if a quiet month leaves it unused. It's what stops a product quietly decaying between big projects.

• You own everything, always

  The code, the cloud accounts, the domains, and every credential are yours from day one, with us added as collaborators. If you ever leave, we run a structured handover to your next team. A client staying out of inconvenience isn't a relationship worth having.

• We adopt software we didn't build

  Orphaned systems get a technical health check, immediate stabilisation of anything dangerous, and documentation written as we go, with an honest assessment up front, including when the right advice is to rebuild rather than keep patching.