Model My Context · Resources

Working smarter and more reliably in government.

AI workflow automation — with you in controlGovernment · EN

Why now?

You know it: growing queues, an inbox full of requests, and people spending more time on admin than on the citizen. Things are moving fast, expectations are rising, and you want to stay ahead — without handing over your data and control. Not adopting AI is not an option; a flawed implementation can damage trust and lawfulness. The question is not whether, but when you prepare your organisation for the future. This document outlines the developments, the processes that lend themselves to it, and how you discover it risk-free in a contained pilot.

Market developments

Labour market in figures (government). Government leads in staff shortages; delivery organisations report sickness absence of around 5 to 8.4% and a scarcity of data analysts, IT staff and heads of information management (Staat van de Uitvoering, 2025). AI support targets the routine work, so your scarce specialists keep time for what only they can do.

Key challenges

The core challenge is clear: improve your service with AI without losing control.

Laws and regulations are stacking up.

Common workflows

The same administrative patterns recur across the sector:

Permit applications and notifications
Objections and appeals
Subsidy applications
Freedom-of-information (Woo) requests
Case intake and citizen correspondence
Procurement and tendering

MMC works modularly. We don't support a single workflow, but build out a growing set of processes step by step — at your own pace, always with human control.

Where AI supports

Within a process, AI supports the repetitive steps while your people decide:

Request / case in

Received

Document analysis

Understand info

Triage & classification

Route

Test against policy & law

Lawfulness

Draft decision

Prepare

Human approval

Judgement

You decide ✓

Handling & archiving

Record

Completed ✓
With you in control. Works alongside your existing systems (no rip-and-replace), with human final control, your data local and under your own control, and no vendor lock-in.

The MMC approach: the pilot

We work modularly, step by step. Together we take one process; you invest minimally, we do the heavy lifting:

  1. Together — we pick one concrete process.
  2. You — an intake of about an hour, plus limited additional information.
  3. We — analyse your process and build a working version.
  4. Together — we test on your own data and discuss the outcome.
  5. You — decide whether and how to proceed. No obligation.

Is your organisation a fit?

Recognition is often a good gauge. This fits if you deal with:

Start a pilot

Discover which workflow in your organisation is the first fit for AI Workflow Automation. Start a pilot and experience how it works on a real process of your own — no obligation, with you in control.
Start a pilot →
Arjan Noordhoek
“A system is not the sum of its parts, but the product of the interactions between them. That's why we don't model the individual steps, but the outcome that matters.”
Arjan Noordhoek — founder & CEO, Model My Context
DownloadWhitepaper as PDF →

Sources: Staat van de Uitvoering (2025) · SER — public-sector labour shortage · AI Act (European Commission) · GDPR (Dutch DPA) · Woo (wetten.overheid.nl) · BIO (Digitale Overheid). Regulatory dates may change; verified June 2026.