The situation
A European energy supplier is moving its entire customer base onto a new platform in groups, and I lead migration support for the programme, a contract running since June 2026. When a group moves, every problem it surfaces has to be owned, and someone has to decide when the next group is safe to go. (The client is unnamed here: the engagement terms restrict it.)
The problem
- The support framework existed on paper but not in practice: no rules for when a group was ready, two overlapping readiness checks, and no way of ranking problems by severity.
- Issues arrived through chat, email and conversation, with no single owner and no shared view of what was open.
- Go-live decisions risked being made on gut feel, by whoever spoke loudest in the meeting.
What I did
- Took ownership of the framework in week one and redesigned it end to end: clear severity rules, one issue board both companies work from, and problems ranked by which group of customers they block.
- Became one of three people with delegated authority to decide go/no-go for each batch of customers going live, and co-decided every gate across the first three groups.
- Logged every decision with the evidence behind it and its outcome (fix now, accept a workaround, or hold), so any go-live can be defended months later.
- Before the fourth group, took six issues marked as blockers back to their owners with one question each: genuinely a blocker, or workable around? The gate opened on evidence, not labels.
- Replaced hand-updated status spreadsheets with one live view read straight from the data, so status meetings stopped spending their first ten minutes agreeing which number was right.
- Automated the issue intake with an AI assistant (its own case study below), and handed the day-to-day running of the process to the wider team by week three.
A migration you can defend is one where every go/no-go decision is written down.