Implementation — Redeploy

What you need to bring

One roster export. CSV or Excel, any column names, any language. If HR can export it, the engine can read it.
One decision on the rulebook. Which capability families move, which locations count, where the score line sits. Defaults are included and sane.
One committee. A small group with the authority to review matches and log overrides. That is the whole organisation chart.

The first week

Day 0

The export

HR pulls the affected team's roster. It lands in the tool as-is — messy headers, Danish fields, abbreviations. The mapping step shows exactly what was recognised and what needs a human glance.

Day 1

The first run

The engine scores the whole roster against every open posting in seconds. Each person gets a match tier and a written reason. No prompts, no waiting.

Days 2–3

The committee pass

The shortlist is reviewed — matches, close fits, and the people with no fit today. Every person is discussed with evidence in front of the committee, not behind a manager's door.

Day 5

The decisions

Assignments are confirmed, overrides are logged on record, and the output — results, unmatched report, written reasons — is ready for the redeployment committee and, if needed, the works council.

Why it stays easy

Nothing to integrate. The job universe comes from the public careers site and refreshes live — postings appear and close on their own. Run the refresh whenever the conversation needs current numbers.
Scales by running, not by building. A thousand people is the same seconds-long run as twenty. There is no pipeline to maintain, no model to retrain.
Governance by default. The rulebook is versioned, overrides are flagged, and every run can be replayed. The audit trail writes itself.
Data stays internal. The roster is a file on your side of the process. No patient data, no third-party database, no cloud dependency you have to justify.

Honest answers to the hard questions

"Will the unions accept it?"

More easily than the alternative. A deterministic rulebook with published weights and per-person written reasons is exactly what a works council can audit. "The system suggested it, here is why, and here is the replay" beats "the manager decided".

"What about bias?"

The engine matches on documented skill, capability and location — not on who knows whom. It cannot inherit a recruiter's blind spots, and because every decision is replayable, any bias claim can be tested in minutes instead of litigated in months.

"What if there are no spots for some people?"

Then they get an honest answer instead of silence: the closest option, with a score, recorded in the output. The business case is built on the people who do move — the calculator shows the effect of any spot/headcount mix.

"Who runs it day to day?"

One person in HR or the redeployment team. Export, run, review, decide. The whole loop is four steps and the tool does the heavy lifting in the middle.

From export to decision in a week.