Her estate was a ten-year-old PHP 5.6 app on a host that had stopped patching. The previous agency wanted to rebuild it over one weekend and switch the DNS on Sunday night. If anything broke, there was no way back.
Migration Software migration services · UK senior-only studio
Software migration services with rehearsed rollback. Zero customer-facing downtime.
Twelve migration paths — every one with a rehearsed rollback, an audited cutover, and a fixed quote at the end of a 5-day audit.
12migration paths shipped
0customer-facing downtime incidents
5 dayfixed-scope migration audit
(Why founders + CTOs sign)
You don’t want a vendor that does this once. You want one that has rehearsed the rollback twelve times.
A migration is the day your business is most exposed. We’ve shipped twelve different migration paths, every one behind a dual-run cutover with the old estate live until you flip the switch.
0
0
£8K
The founder this page is forPHP 5.6 · 27 plugins · no rollback plan
Nadia’s last quote read “big bang weekend cutover.” No rollback. No dual-run.
That’s not a migration, that’s a coin flip with your revenue. We don’t do big-bang weekends. We build the target in parallel, shadow-read live traffic, then flip the write path behind a feature flag. The old estate stays live until you’re sure.
Every cutover we ship keeps a 30-day rollback window open. If a number looks wrong on day 12, we flip back in minutes, not weeks. This page is for founders who’d like to migrate without holding their breath.
The migration fear. The day-1 answer.
Each one is a single decision we make on day one of the build, before a line of cutover code is written.
- 01
“The last quote was a Sunday-night DNS flip with no way back.”
Dual-run, shadow-read, flag-flip cutover. The old estate stays live. We shadow-read live traffic against the new one, then flip the write path behind a flag. No weekend coin flip.
- 02
“If something looks wrong after cutover, are we stuck?”
30-day rollback window, rehearsed. We rehearse the rollback before cutover and keep a 30-day window open. If a number looks wrong on day 12, we flip back in minutes.
- 03
“We can’t lose a single order or customer record.”
Reconciliation diff on every table. Logical replication for databases, row-count and checksum reconciliation before and after. Encrypted fields ported with key rotation, decryption tested.
- 04
“We migrated once before and lost half our rankings.”
1:1 redirect map, validated pre-cutover. Every URL inventoried and mapped, validated against Search Console traffic before we flip. Rankings come across with you, not after you.
- 05
“We can’t stop shipping features for six months.”
Strangler-fig, surface by surface. We move one surface at a time behind a proxy. The old app keeps earning and your team keeps shipping while the migration runs underneath.
- 06
“Our UK and EU customers need their data to stay in region.”
UK-resident by design, proven in writing. On-premise to cloud with eu-west-2 or London regions, residency evidenced for the questionnaire, not just claimed in the proposal.
- 07
“Nobody here actually knows everything the old system does.”
5-day audit, then a fixed quote. Two senior engineers map the estate, cron jobs, integrations and dark corners in 5 days. You get a risk matrix and a fixed price before any build.
- 08
“What if your engineer quits halfway through our migration?”
Two seniors paired, ADR same day. Every migration runs with two senior engineers, not one. Decisions go into an ADR the same day. Two engineers have left in seven years, both handovers inside 48 hours.
Legacy modernization UK · senior-only studio
Four reasons product teams finally move off the legacy stack
None of these are taste. Each one is a cost you’re already paying, in slower delivery, a smaller hiring pool, a bigger blast radius, or a failed security questionnaire.
You can hire for the new stack
A PHP 5.6 or CodeIgniter 2 estate is a hiring dead end. Move to Laravel, MERN or modern PHP and your next senior interviews fluent on day one.
Delivery speeds up, not slows
Strangler-fig means features keep shipping while the migration runs underneath. You’re not frozen for six months waiting on a rewrite.
Your cloud bill stops creeping
On-premise to cloud and serverless migration trade fixed idle cost for scale-to-zero. You pay for the traffic you actually serve.
You pass the security questionnaire
An unpatched legacy host fails diligence. A modern, observable, UK-resident estate answers the 60-page questionnaire in writing.
What we migrate onto, and the tools that move it.
The target depends on your estate. The tooling is constant: replication, reconciliation, redirect maps, and observability wired in before the first cutover.
T1
Targets we migrate onto
modern stacksReact + Next.jsNode.jsMongoDBPostgreSQLLaravelSanity / ContentfulGraphQL (Apollo / Hasura)FlutterShopifyTypeScript
T2
The tooling that moves the data
migration kitLogical replicationpgloaderAWS DMSDrupal Migrate APIStrangler-fig proxyReconciliation diffsFeature flags
T3
The infrastructure layer
AWS-defaultAWSLambda + FargateEventBridgeDynamoDBS3 + CloudFrontRDS / AuroraDockerTerraformDatadog + Sentry
Software migration services · the paths we ship
The migrations we ship most
Each one behind a dual-run cutover, a rehearsed rollback, and a fixed quote at the end of a 5-day audit.
12+
Migration paths shipped
0
Customer-facing downtime
“They built the new estate alongside the old one and shadow-ran it for two weeks. We flipped on a Tuesday afternoon and nobody outside the team noticed.”
001
WordPress + Drupal to headless
WordPress and a wall of plugins, or Drupal 7, to Next.js + Sanity. Editors keep their workflow, the front end gets fast, redirects validated first.
002
Legacy PHP to modern
Laravel, CodeIgniter, CakePHP or Yii to Laravel 10/11 or MERN, strangler-fig style. The old app keeps earning while we move it off a stack you can’t hire for.
003
Database + API migration
MySQL to PostgreSQL via logical replication, REST to GraphQL with Apollo or Hasura. Zero downtime, reverse-replication rollback ready to fire.
004
On-premise + serverless
Datacentre to AWS, GCP or Azure with UK data residency proven, or EC2 and VMs to Lambda + Fargate. You stop paying for idle, scale-to-zero kicks in.
005
Native to cross-platform + Shopify
Swift and Kotlin to Flutter or React Native, one team and two stores. WooCommerce to Shopify with products, orders, customers and SEO carried across 1:1.
What teams say after a safe cutover
From the founders, COOs, and CTOs who moved off the legacy stack without holding their breath.
100%
would refer us
They built the new estate alongside the old one and shadow-ran it for two weeks. We flipped on a Tuesday afternoon and nobody outside the team noticed.
What CTOs actually ask before signing
Pain-first, soft-second.
We don’t do big-bang weekend cutovers. We build the target in parallel, shadow-read live traffic against it, reconcile every record, then flip the write path behind a feature flag. The old estate stays live until you’re sure. That’s how every one of our software migration services keeps zero customer-facing downtime.
We rehearse the rollback before we cut over and hold a 30-day rollback window open afterwards. For databases that means reverse replication kept running. If something looks wrong on day 12, we flip back in minutes, not weeks, and diagnose with the old estate still serving traffic.
The audit is fixed at £8K for 5 days. After that you get a fixed quote for the migration itself, never a number before we’ve seen the estate. Scope is driven by estate size, integrations, data volume and how much custom logic is buried in the old system. You’re never more than a few days at risk before you see real work.
No, not if it’s planned. Every URL is inventoried and mapped 1:1, then the redirect map is validated against your Search Console traffic before cutover. We carry rankings across with you. The migrations where rankings collapse are the ones where nobody built the redirect map first.
Yes. We use the strangler-fig pattern, moving one surface at a time behind a proxy. The old app keeps earning and your team keeps shipping while the migration runs underneath. You’re not frozen for six months waiting on a rewrite that may or may not land.
30-day walk-away both ways. Milestone billing. UK VAT registered, listed on Companies House, shipping since 2019. The audit is fixed at £8K, so you’re never more than a few days at risk before you see real, useful work in your hands.
We default to AWS eu-west-2 (London) for UK data and pin EU data to an EU region where it’s required. Residency is evidenced for your security questionnaire, not just claimed in a proposal. On-premise to cloud migrations are where this matters most, and it’s mapped in the audit.
Every migration runs with two senior engineers paired, not one. Every decision goes into an ADR the same day. Mohit reviews every PR. If one engineer leaves, the other has full context the next morning. Two engineers have left mid-project in seven years; both handovers were inside 48 hours.
Most likely, yes. The twelve paths on this page are the ones we ship most, but the audit maps any estate. If yours is genuinely outside what we should take on, we’ll tell you that in the reply rather than after the deposit. Send the brief and we’ll be straight with you inside 24 hours.
In context
The surface you hand over.
software migration services, in context — the dashboards, flows and components your team actually ships, reviews and maintains.
Twelve paths. One safe cutover. Start with the audit.
Send a 5-line brief: current estate, target stack, the deadline you’re working to. Mohit replies inside 24 hours with availability and the next 5-day audit slot.
What happens after the email lands
- < 24h
A personal reply.
Availability and the next 5-day audit slot, straight to your inbox.
- Week 1
Audit week begins.
Risk matrix, target architecture, data plan, rollback plan, fixed quote.
- Cutover
Zero downtime.
Dual-run, shadow-read, flip the write path. Rollback rehearsed.