Skip to content

Migration

Your system in the cloud. In months, not years.

A legacy migration is mostly volume: screens, reports, data, and integrations. Volume is what a swarm is for. One principal directs it. The old system stays live until the new one matches.

Why migrations take years.

Not because the work is hard. Because a team is the unit, and a team has a ceiling.

Volume

Hundreds of screens and reports, each rebuilt by hand. The estimate is a headcount multiplied by a year.

Handoffs

Analysts, developers, testers, and data people work in sequence. Every handoff is a week.

The switch

After all that, the cutover is a weekend nobody has rehearsed, with no written way back.

While it runs

What you can see at any point.

Revenue

Revenue stays on the old path

  1. 01 Beside The new system runs on real data. Customers do not move yet.
  2. 02 Proof Totals match, line by line. Finance signs the comparison.
  3. 03 Cutover You choose the hour traffic moves.
The old system still takes traffic until finance signs that the numbers match.

The work

How it runs, in order.

01 First plan

Inventory and fixed price

Every screen, report, integration, and data set is counted. You get a price for that scope and a date. Both are written.

02 Parallel build

The swarm rebuilds it

Screens, reports, services, and data migration run at the same time. The principal reviews. Tests are written with the code.

03 Dual run

Both systems, real data

The new system runs beside the old one. Numbers are reconciled line by line until finance signs the match.

04 Cutover

Switch, then turn off the old one

You choose the hour. The rollback is rehearsed. Your team accepts the files. Our access is removed.

01 Parts

One part at a time. Nothing you cannot undo.

We move the system in parts, starting with the one that would break month-end. If a part fails, the rest stays up. The rollback for each part is written before it switches.

In parts

Pause one part. Keep the rest running.

  1. 01 First We start with the part that would break month-end.
  2. 02 Beside The old path still takes customer traffic.
  3. 03 Cutover That part switches only after someone signs the match.
One paused part does not take the rest of the system down.
02 Proof

The match is signed, not shown on a dashboard.

The swarm reconciles the new system against the old one on real data, every day, at a volume no team could sustain. Finance signs the comparison. Then you switch that part.

Revenue

Revenue stays on the old path

  1. 01 Beside The new system runs on real data. Customers do not move yet.
  2. 02 Proof Totals match, line by line. Finance signs the comparison.
  3. 03 Cutover You choose the hour traffic moves.
Customers stay on the old path while the new one proves itself.
03 Go-live

We make it safe. You decide when.

The principal writes the switch criteria and the rollback and holds the line if a check fails. You choose the hour live traffic moves.

Who holds it

Delivery risk has one owner

  1. 01 Hours Many vendors, many hands. You own every miss.
  2. 02 UNIT01 One principal owns the result until you accept.
  3. 03 Live You decide when traffic moves. We make sure you can.
Moving live traffic is your decision. Making it safe is ours.

In the first plan

What is written before work starts.

The first plan names each of these. It is a document you can accept, edit, or decline.

Scope

Counted, not estimated

Every screen, report, integration, and data set is listed. The price is fixed to that list.

Principal

One named person

The engineer in charge is named in the contract, with a deputy. You meet them before you sign.

Match

Reconciliation as a term

Behaviour and numbers must match the old system before any part switches. Finance signs.

Rollback

Written and rehearsed

Time to reverse, data implications, and who calls it. Rehearsed before the cutover, not after.

Files

In your repository

Code, tests, runbooks, monitoring, and open risks. There before acceptance.

After

Warranty and access

A warranty period on the agreed work. Our access removed at close. Both written.

Questions we get asked

Common questions.

01

How can it be this fast?

A migration is mostly volume. A swarm of agents rebuilds screens, reports, and data mappings in parallel, and the principal reviews. The volume that costs a team a year is the part that costs us the least.

02

How can the price be fixed?

Our cost is not headcount. We count the scope, price it, and hold it. If the scope changes, the price changes in writing. Nothing else moves it.

03

Do you rewrite everything?

No. Move, wrap, or retire. The first plan says which, screen by screen.

04

Can the old system stay up?

Yes. Running both is the default on revenue paths until the numbers match.

05

What about our systems integrator?

They can keep a part. We will say in week one if the split is unsafe, so it never becomes a fight.

06

Where does the data live?

In your accounts, under your residency rules. We do not become the system of record.

Other work

The same way of working, on a different system.

Next step

Tell us what must go live, and when.

We come back with a first plan: what done means in your words, what it costs, and when. Then it is your call.