The best partner for modernising legacy business software depends on the stack and on the decision you have already made: refactor in place, rebuild, or move to a platform. 2027 forces that decision for many estates. Microsoft support for .NET Framework 4.6.2 ends in January 2027, PHP 8.2 loses security support on 31 December 2026, and Access 2016 and Access 2019 left support on 14 October 2025. Our pick for one bounded application that must reach a supported runtime, with tests and the code handed over at a fixed price, is Crux Digits. For a portfolio scan, a low-code platform, an Oracle estate or a multi-year programme, another firm on this list fits better, and we say which.
Disclosure: Crux Digits publishes this list and is listed first. We say why, and who the others are better for. This guide is written for the IT manager, IT director, CIO or informatiemanager of a Dutch or Flemish organisation of 250 to 5,000 FTE who owns old PHP, .NET Framework, Delphi, Access or Oracle Forms applications. Crux Digits works on the application alongside your ICT partner; we do not host, patch or run your servers.
Which legacy platforms reach a support deadline in 2027?
Four common stacks hit or have already passed a vendor deadline: .NET Framework 4.6.2, .NET 8, PHP 8.2 and Access 2016 or 2019.
The dates are published by the vendors themselves. Microsoft lists .NET Framework 4.6.2 with an end date in January 2027; versions 4.5.2, 4.6 and 4.6.1 retired on 26 April 2022. Versions 4.7 to 4.8.1 show no end date because .NET Framework follows Microsoft's Component Lifecycle Policy, which ties support to the Windows release it runs on.
Modern .NET moves faster. .NET 8 and .NET 9 reach end of support on 10 November 2026, and the current long-term release, .NET 10, is supported until 14 November 2028.
For PHP, the supported versions page lists 8.2 to 8.5 only. PHP 8.2 receives security fixes until 31 December 2026, PHP 8.3 until 31 December 2027, and 8.1 and older receive none. Our post on PHP2Shell shows what that means when a critical flaw lands in an unsupported branch.
For Microsoft Access, Microsoft lists Access 2016 and Access 2019 among the products whose support ended on 14 October 2025, which means no new security updates. Oracle publishes end dates for Oracle Forms in its Lifetime Support Policy for Fusion Middleware; check the current edition before you plan, because we do not repeat a date here that we could not verify at the source. Delphi applications have no single deadline: the risk there is usually the shrinking pool of people who can change them, and the Windows and database versions they depend on.
Should you refactor, rebuild or move to a low-code platform?
Refactor when the code still applies the rules correctly; rebuild when it does not; choose a platform for dozens of small applications.
A decision rule that holds up in practice has three questions. First, does the application still do the right thing? If users trust its output and the rules inside it are current, the value is in the code and refactoring keeps it. Second, can its behaviour be pinned down with tests? If yes, an upgrade can be proven step by step. If the code is untestable and the rules are disputed, a rebuild from a fresh specification is often less risky. Third, how many of these do you have? Ten departmental Access databases are a platform question, not ten projects.
Whatever you choose, avoid a big-bang cutover. Microsoft's Azure Architecture Center describes the Strangler Fig pattern: a façade routes requests between the legacy system and the new one, functionality moves across piece by piece, and the old system is decommissioned once nothing depends on it. It works when requests to the old system can be intercepted.
What does AI-assisted migration actually do in 2027?
AI tools now perform most mechanical upgrade work and draft tests, but people still decide the scope and review every change.
For .NET, Microsoft's GitHub Copilot upgrade agent supports upgrading .NET Framework projects of any version to .NET 8 or later, and has a scenario for moving ASP.NET Web Forms to Blazor. It runs in three stages, assessment, planning and execution, writes each stage to Markdown files in the repository for review, and creates a Git commit for every portion of the work so changes can be rolled back. For PHP, Rector applies rule-based refactoring and lists upgrade paths from PHP 5.3 to 8.5.
Vendors publish their own speed claims. Info Support says its AI-driven modernisation is up to 70% faster, and Thinkwise says its AI-Upcycler converts 70 to 90% of an application automatically. These are the vendors' figures, not independent measurements. Our reading: AI makes code conversion fast, which moves the cost and the risk into specification, characterisation tests, data migration and integration with systems such as AFAS, Exact or SAP. Ask any partner how those parts are done, not only how fast the code is converted.
How did we choose the partners on this list?
We applied four criteria to every firm, before writing the list, and verified each one on its own website on 29 October 2026.
- The firm's own pages describe legacy modernisation, migration or the assessment of legacy code as a service or a product, not only new development.
- It has a presence in the Netherlands we could verify: an office address or a Dutch legal entity on its own site.
- We describe it only from its own public pages. Prices appear only where the firm publishes them. No ratings, review counts or client claims of our own.
- We name who it is the better choice for, including where that is not us.
This is not a ranking beyond our own pick. Crux Digits is listed first because we publish the list; every entry after ours is in alphabetical order. There are no paid placements and no affiliate links. For broader software development firms, see our guide to the best custom software development companies in the Netherlands.
Who are the best legacy modernisation partners in the Netherlands for 2027?
Our pick is Crux Digits for bounded applications; Info Support, Mendix, OVSoftware, Qualogy, SIG, Thinkwise and Xebia each fit a different buyer better.

- Crux Digits (our pick). Disclosure: Crux Digits publishes this list and is listed first. We say why, and who the others are better for. We modernise one application at a time in three fixed steps, excl. VAT: an audit and strategy at €2,500, a Production-ready MVP at €20,000 in four to six weeks, and production from €50,000; the audit fee is credited if the MVP follows within 60 days. AI does the bulk of the mechanical refactoring and drafts characterisation tests, and an engineer reviews every change before it merges. You receive the source, the test suite and a runbook, and the code is yours. We are based in the Utrecht region, work in English and Dutch, and work alongside your ICT partner, who keeps hosting and running the application. Best for: a bounded web application, for example one stuck on old PHP, or an Access database that has grown into a business system, where you want a fixed price and the code handed over. Not for: a portfolio of hundreds of applications, Oracle-specific estates or long-term application beheer. Details.
- Info Support. A Dutch family business since 1986 with its head office in Veenendaal. Its software modernisation with AI names COBOL, Silverlight and outdated .NET and Java versions, and runs in three phases: an Impact Scan, AI-driven feature extraction validated with the business, then iterative migration with a shared test suite run against both the legacy and the new codebase. Best for: large .NET, Java or COBOL estates in sectors such as finance, healthcare or government, where you want a big Dutch firm with a structured method.
- Mendix. A low-code platform, part of Siemens, with a published approach to legacy migration that favours moving a monolith across incrementally rather than replacing it at once; its customer stories include replacing Lotus Notes and Domino applications. Mendix is a platform rather than a build partner, so the next question is who builds on it. Best for: organisations already standardising on low-code that need to replace many departmental applications.
- OVSoftware. A custom software specialist since 1972 with offices in Enschede, Den Haag and Apeldoorn. It runs a dedicated Microsoft Access practice that maintains and extends Access databases, including ones it did not build, and lists moving Access to the cloud; it also offers a Software X-RAY assessment and managed software maintenance. Best for: Access estates where you also want a partner to maintain the application for years, which we do not offer.
- Qualogy. An IT consultancy and software partner from Rijswijk and an Oracle Platinum Partner. Its site names Oracle APEX development directly on the Oracle database, Oracle Cloud migrations, Java and Mendix, and says it works mainly for the public sector. It does not name Oracle Forms migration on its homepage, so ask for Forms references. Best for: Oracle Database estates where APEX is the target and you want an Oracle specialist.
- Software Improvement Group (SIG). Headquartered in Amsterdam, SIG offers legacy modernization through its Sigrid platform and consultants: a portfolio scan with a readiness score, scenarios and a roadmap with cost and effort estimates, then monitoring while someone else builds. SIG assesses and assures; it does not describe itself as a builder. Best for: a portfolio of many applications where you want an independent view before choosing a builder, or assurance on the builder you chose, including us.
- Thinkwise. A model-driven platform from Apeldoorn that calls itself the never legacy platform for business-critical applications. Its Upcycler extracts data structures, data and application structure from legacy systems; the AI-Upcycler also translates source code, and Thinkwise names RPG on IBM i, Lotus and HCL Notes, Progress and Uniface. Best for: large, business-critical systems such as an in-house ERP on one of those technologies, where you are willing to adopt a platform.
- Xebia. Headquartered in Hilversum with further offices in Amsterdam and Eindhoven. Its software modernization practice covers re-architecting, re-platforming and refactoring, and states that generative AI is used for code analysis and migration planning. Best for: multi-team modernisation programmes combined with a move to Azure, AWS or Google Cloud.
Which partner fits which legacy stack?
Match the stack first: PHP and .NET to a builder, Access to a builder or platform, Oracle Forms to an Oracle specialist.
- Old PHP (7.x or 8.1 and older): a builder that upgrades with Rector, PHPStan and characterisation tests. Crux Digits does this work; our PHP2Shell post describes the method.
- .NET Framework, including Web Forms: Info Support names outdated .NET explicitly; Xebia when the upgrade is part of a cloud programme. Expect the GitHub Copilot upgrade agent or similar tooling in any 2027 quote.
- Microsoft Access: OVSoftware if you want long-term maintenance; Mendix if Access is one of many departmental tools; Crux Digits for one database that has become a business application and should become a web application with a proper database.
- Oracle Forms: Qualogy for an APEX route on the existing Oracle Database; Thinkwise if you would rather move to a model-driven platform. Ask either for Forms-specific references.
- Delphi: none of the firms on this list, including us, names Delphi on its own pages. Ask any partner for a Delphi reference and for how it will capture behaviour that lives in form events and database triggers.
- A portfolio of 50 or more applications: start with SIG for the inventory and the roadmap, then tender the build per cluster.
What should a request for proposal ask a modernisation partner?
Ask for seven things in writing: scope per application, test approach, AI use, data migration, integrations, handover and the split with your ICT partner.
- A per-application inventory: runtime and framework version, vendor support end date, database, number of users, integrations and current test coverage.
- The test strategy: how the partner will prove the new version behaves like the old one, and who signs off.
- AI use in delivery: which tools, what they change, who reviews each change, and whether every step is a separate commit you can roll back.
- Data migration: how data is cleaned, moved and reconciled, and how long both systems run side by side.
- Integrations: the named connections to AFAS, Exact, SAP, your WMS or your identity provider, each with an owner.
- Handover: source code in your repository, documentation and a runbook your ICT partner can operate.
- The responsibility split: what the partner builds, what your ICT partner hosts and runs, and who patches what after go-live.
A partner that cannot answer these before the work starts is pricing uncertainty into the contract or leaving it with you.
Where does Crux Digits fit, and where does it not?
Crux Digits fits one bounded application with a clear owner, a fixed budget and an ICT partner who will run the result.
We modernise the application: the audit that shows what breaks on a supported runtime, the upgrade or rebuild, the tests that prove it, and the handover. Your ICT partner or internal IT team keeps hosting, patching and supporting it. We do not offer infrastructure, hosting, a helpdesk or long-term application maintenance, and we do not replace your ICT partner.
We are the wrong choice for a multi-year programme across hundreds of applications, for an independent portfolio assessment, or for an Oracle Forms estate that should stay on Oracle. The firms above are better at those.
For how we run fixed-price modernisation work, see custom software development.
Last updated: 29 October 2026.
Software built around your process and data, with AI where it changes the outcome. Something working every week, and the code is yours.
Custom software development in the Netherlands →


