The situation
A portfolio that had grown product by product: six separate customer-facing interfaces, three products without one, more than eight codebases. Each product had its own way of working, its own design, its own release pattern. The user interfaces were obsolete. The cost of keeping it all running was high, customer satisfaction was under pressure, real innovation was not happening, and several of the codebases were unstable.
Everyone could see the problem. What nobody had was a decision that would survive the organisation.
What was on the table
A front-end redesign. Put a new skin on the existing products. Fast, visible, and worthless: it would have changed nothing about the eight codebases, the release patterns or the cost.
A complete rebuild. Replace everything at once. Rejected on cost and, more importantly, on knowledge: the understanding of what the products actually did for municipalities lived in the existing code and the people around it, and a big-bang rebuild would have thrown that away.
A wind-down of the legacy with a successor to follow. In Dutch, a sterfhuisconstructie: let the old products run out while something new is built elsewhere. Rejected because attrition on the old products would be fast and the replacement was not in place. The customers would have been left in the gap.
The decision
Build the iConnect Regie Platform: one platform with shared services, and the products as modules on it, migrated product by product rather than all at once. Turn what had been a proposition on slides into an actionable, value-adding product with a roadmap, a team and a first customer-validated result.
As Strategic Product Manager for Information and AI I own the direction of that platform and the eight product and innovation lines around it, across a portfolio that serves 150 municipalities, provinces and water boards.
The organisational part
The technical decision was the easy half. Resistance was general, across teams and operating levels, and it was earned: earlier attempts had not landed, and people had learned to expect that this one would not either.
The answer was not to persuade the whole organisation first. It was to build a new team with new hires, protect its momentum, and let that team help the rest of the organisation navigate its own obstacles and reluctance with something working to point at. Leading that change is as much the job as the platform is.
Status
| Milestone | State |
|---|---|
| Ideation to first result | Done; validated with customers |
| First production customers | Planned Q2 2027 |
| Ideation to production | Eighteen months |
Evidence grade: pilot. The result has been validated with customers; it is not yet in production, and I will update this page when it is.
What it taught me
The alternative you reject tells people more than the one you choose. Saying no to the redesign, the rebuild and the wind-down, each with its reason, is what made the platform decision credible.
Resistance is information. Years of attempts that did not land is a fact about the organisation, not about the people. A new team with momentum changes the fact.
Platform first, AI second. Every AI ambition in this portfolio runs on this platform. Without one platform, there is no governed place for agents to run. That is why this decision sits underneath everything else on this site.