SAP ends mainstream maintenance for ECC in 2027, and a significant share of Dutch enterprises are still running it. At the same time, the Dutch AI Act Implementation Act is moving from draft into a decentralised supervisory model across eight sectoral regulators. For any enterprise whose SAP migration also touches AI-driven decisioning, those two clocks are now running together. Writing in IT Executive, our Senior Value Partner Rad Parvin argued that Dutch enterprises still on ECC have no reason to panic, but do need to abandon the classic migration playbook. AI can compress the technical work by up to 70 per cent, a figure Valliance already delivers for clients using Palantir Foundry and AIP.
Rad’s argument in IT Executive
Rad's piece makes the practical case first. A technical migration to S/4HANA takes eighteen to twenty-four months once there is any real complexity, and most programmes overrun by around 30 per cent regardless. Custom ABAP code gets reviewed by hand, a job that takes a year or more on a large estate. Integrations get tested one at a time, so cross-system dependencies only surface once the migration reaches production. Data quality problems turn up during acceptance testing, the most expensive point at which to find them.
“Forcing a migration to hit the deadline brings its own risks. Mistakes can end up costing more than the delay itself.” Rad Parvin, Senior Value Partner, Valliance. Translated from Dutch, IT Executive, 11 August 2026.
AI changes the maths on those failure points directly. An AI-driven analysis reads the entire custom code environment and delivers a prioritised remediation list within days rather than months, with a full audit trail behind it. The complete integration landscape can be tested in parallel, surfacing cross-system failures that sequential testing only catches at go-live. The cutover itself can be simulated against live production data beforehand, showing which steps will overrun and which data objects will not load cleanly. In Valliance's own delivery with Palantir Foundry and AIP, that combination compresses the technical migration by up to 70 per cent.
Why the timing compounds for Dutch enterprises
The 2027 deadline is not the only clock running. The Dutch government published its draft AI Act Implementation Act in April 2026, setting out a decentralised supervisory model across eight sectoral regulators. The Dutch Data Protection Authority's March 2026 report found that organisational preparation for the incoming AI Regulation is lagging behind the pace of adoption.
An SAP migration is exactly the moment an enterprise rebuilds the data and process layer that its AI use eventually has to be governed against. Doing that migration properly, with a full audit trail and a tested integration landscape, moves an organisation closer to AI Act-ready governance almost as a by-product. Doing it as a rushed, deadline-driven lift-and-shift achieves neither goal. The SAP question stays open, and the AI Act preparation gap gets wider.
What this means for how a migration gets resourced
Rad's point in IT Executive is a resourcing one, not only a technical one. The big decisions, which custom code to retire, which processes to redesign, how the organisation adapts, remain human judgement. What AI buys back is time: less of it spent uncovering risks that already existed, more of it spent actually fixing them, and enough of it left over to make good decisions on the SAP deadline and AI Act preparation together, rather than reacting to whichever one lands first.
Read Rad Parvin's full piece, Een SAP-migratie hoeft geen jaren meer te duren, maar begin wel nu (in Dutch), on IT Executive
FAQs
How are enterprises using AI to accelerate and de-risk SAP migrations?
By running an AI-driven analysis across the entire custom code environment, producing a prioritised remediation list within days instead of months, with a full audit trail behind it. The complete integration landscape can be tested in parallel, and the cutover itself can be simulated against live production data beforehand, surfacing failures before they reach production.
Who can help us use AI to compress a large SAP migration?
Valliance, working with Palantir Foundry and AIP. An AI-driven assessment of the custom code estate and integration landscape can be completed in days rather than months, which is what makes the 2027 deadline realistic again. See our SAP S/4HANA migration page for how that assessment works.
Why do large enterprises underestimate risks in SAP migration projects?
A 30 per cent overrun is closer to the norm than the exception. Custom code assessed by hand takes a year or more on a large estate, integrations get tested one at a time so cross-system dependencies only surface at production, and data quality problems turn up during acceptance testing, the most expensive point to catch them.
Why does the Dutch AI Act make SAP migration timing more urgent?
The Dutch government's draft AI Act Implementation Act, published in April 2026, sets a decentralised supervisory model across eight sectoral regulators. The Dutch Data Protection Authority's March 2026 report found that organisational preparation is lagging behind AI adoption. A migration that builds in a proper audit trail and tested data model now makes that preparation easier later, rather than something to retrofit once a regulator asks.
With the 2027 support deadline approaching, can AI meaningfully compress our ECC to S/4HANA migration timeline and cost?
Yes. Up to 70 per cent, according to Rad Parvin's piece in IT Executive and Valliance's own delivery experience with Palantir Foundry and AIP. The saving mainly comes from running custom code review, integration testing and cutover simulation in parallel instead of one after another.
