I translate business goals into technical reality — so your plans actually ship.

Regulatory reporting, system architecture and AI adoption are three places where business language and technical detail collide. I sit in the middle and make sure what you need is what gets built: a reporting obligation you actually understand, a working integration, an AI plan that holds up.

Book an intro call

Independent consultant since 2003 · Co-founder of Taxxor · Geldrop, the Netherlands

Reporting Systems AI JOES Shipped

Three ways I can help

One translation skill, applied to three domains where business language and technical reality collide.

Reporting Process Optimization

For finance and reporting teams stuck in manual work

Reporting periods that run on copy-paste, spreadsheet juggling and late nights are a process problem, not a people problem. I analyze how the work actually flows and apply technology where it pays off — so the numbers are right and the deadline isn't a crisis.

  • Process analysis: where time is lost, where errors creep in, what to automate first
  • Connecting reporting to your source systems — ERP and consolidation included (SAP, LucaNet)
  • Translation between finance, auditors, IT and your software supplier
Need reporting software rather than advice? That's Taxxor's territory — I'll gladly introduce you. My consultancy covers the process, tagging and integration questions around it.

Proof: co-founder of Taxxor; reporting-process work from SAP BW/BCS programs in a multinational to disclosure automation for medium and large corporations.

Architecture & Integration

For organizations that need a senior architect, not a full-time hire

As a fractional software architect I bring 25+ years of design and integration judgment to your team — useful when systems multiply faster than the people who have to connect them.

  • Fractional architect engagements, 1–3 days per week
  • Architecture and integration roadmaps, API and data-exchange strategy
  • Stepping into complex, partly-documented systems: mapping what's actually there and planning the way out

Proof: 25+ years as developer and architect — from CMS platforms at the design studio of a global electronics multinational to recently leading a development team through an AI-assisted legacy-platform migration.

Pragmatic AI Consultancy

For teams that want AI in their processes — without the hype

I help you decide where AI genuinely helps, which tools fit, and how to adopt them responsibly. The deliverable is judgment, not slideware.

  • AI process scan: where AI helps, and where it doesn't
  • Local / on-premise AI advisory when data is too sensitive for the cloud
  • Tool selection, implementation guidance, GDPR and EU AI Act sanity check

Proof: hands-on with cloud and local AI (Claude, Ollama, LM Studio); recently used AI tooling to map a legacy application and generate its migration plan. No tool to sell, no vendor commissions.

Book an intro call

How I work

The same four steps, whether the subject is a reporting obligation, a system landscape or an AI plan.

Understand the actual goal

Not the ticket, the goal. I start with your business objective and constraints, so we solve the right problem instead of implementing technology for its own sake.

Translate it into specifications everyone signs off on

Business needs, design intent and technical constraints land in one document that finance, designers and developers all read the same way. This is where misalignment dies.

Price on value, not hours

Engagements are structured around the result you need — a working integration, a decision you can defend, a plan your team can execute — not a timesheet.

Implement, then hand over

Whether I build it or guide your team, knowledge transfer is part of the job. You should be able to run and evolve the result without me.

Cases

Four examples of the same pattern: a business problem stuck on technical detail, translated into something that shipped.

Building a disclosure platform for regulated filings

Problem
Digital filing mandates (ESEF, SEC, SBR) left reporting teams stuck between regulator requirements written for technologists and tooling built for accountants.
Solution
Co-founded Taxxor and helped build its disclosure management software around one idea: a single source of truth feeding every output. Figures flow in from systems like SAP S/4, SAP BW and LucaNet; reports flow out as XBRL, designed PDF, web and Word/Excel — map once, report infinitely.
Result
In production at listed companies, housing corporations and education institutions — supporting every ESEF filing season since the 2021 mandate, alongside SEC 20-F, Dutch SBR and ESG reporting from the same source. The reason my advice comes from building this technology, not reading about it.

Keeping design intent alive through implementation

Problem
At the in-house design studio of a global electronics multinational, creative vision arrived as pixels, developers needed specifications, and intent got lost in between.
Solution
Worked alongside the visual designers as the live translator — converting design vision into buildable requirements, and feeding technical constraints back before they became surprises.
Result
Designs shipped as designed — from the global brand-book intranet and extranet to the annual report's online channels. One first: readers composing their own PDF of the annual report online, a designers' concept I translated into a working application that still felt completely on-brand.

Landing a SAP reporting change across business and IT

Problem
A multinational upgraded its SAP BW/BCS inter-company cost allocation — overhead, brand fees and licenses redistributed by an allocation key agreed with the tax authorities. The change had to land with business, finance and IT at once.
Solution
Coordinated test efforts across business and IT, wrote the test scripts, produced usable documentation and training material, and ran the knowledge transfer between the project team and internal IT.
Result
The updated allocation solution was adopted by the people who have to run it — tested by the business, documented for IT, no orphaned project deliverables.

Documenting and future-proofing a legacy platform with AI-assisted analysis

Problem
A digital agency group manages business-critical software for one of the Netherlands' largest home-improvement retailers — on an ageing legacy platform that was only partly documented. With frequent turnover in the supporting team onboarding had become a real bottleneck and a business risk.
Solution
Working with the team of developers and business analysts, used AI tooling to analyze the legacy code and turn it into living documentation the team could query directly — then defined a short- and medium-term strategy to stabilize the platform and gradually modernize it, benefiting all parties involved, including the retailer's customers.
Result
New developers now onboard against documentation instead of guesswork, the platform stays stable while it evolves, and the client has a concrete plan to modernize on their own timeline — a case of AI applied where it pays off, not where it's fashionable.
Book an intro call

Working together

Three ways in, from small to structural. No public price list — the first conversation is free and tells us both whether there's a fit.

Step 1 · Smallest commitment

Advisory session

A focused session on one question — a filing approach, an architecture decision, an AI plan. You leave with a written recommendation you can act on, with or without me.

Step 2 · Defined scope

Project engagement

A defined problem with a defined deliverable: a reporting readiness review, an integration roadmap, an AI adoption plan. Priced on the result, not the hours.

Step 3 · Structural

Fractional architect

One to two days a week as your software architect — senior judgment embedded in the team, for as long as the gap exists.

Book an intro call

About Johan

Johan Thijs, technical-business translator

I started building internet applications at the medical systems division of a global electronics multinational in 1999, moved to its corporate design studio as senior application developer, and have run JOES Internet Services as an independent consultancy since 2003. Since 2017 I'm also managing director and co-founder of Taxxor, a disclosure management software company.

Most of that career is one skill in different costumes: translating between people who describe a problem in business terms and people who have to solve it in technical terms. At the design studio that meant turning creative vision into specifications that survived implementation. At Taxxor it means building structured reporting technology (XBRL, iXBRL) against ESMA, CSRD and SEC IFRS standards. In between sit two decades of content management, system integration and application architecture — including SAP BW/BCS reporting work in a multinational, coordinating business, finance and IT.

JOES Internet Services is where that experience is available directly: on your reporting processes, as a fractional architect in your team, or as a sparring partner for your AI plans. I listen, translate it into something everyone can sign off on, and then build it or guide your team while it's built.

Frequently asked questions

What does optimizing a reporting process look like in practice?

We map how the report actually comes together: which numbers come from which systems, where people copy-paste, where versions diverge and deadlines slip. Then we fix the expensive parts — automating data collection and validation, connecting your ERP or consolidation system to your reporting tools, removing the manual steps where errors are born. Technology carries the load, but the deliverable is a calmer reporting period, not just new software.

What is iXBRL, in plain language?

Inline XBRL is a normal, human-readable report (a web page) with machine-readable labels hidden inside it. A person reads the annual report; software reads the same file and extracts every tagged number with its definition, period and currency. One document, two audiences — which is exactly why regulators standardized on it.

Does CSRD change my digital reporting obligations?

Yes. CSRD sustainability disclosures are designed to be digitally tagged like financial statements, using an ESRS taxonomy. If ESEF taught reporting teams to tag their financials, CSRD extends that discipline to sustainability data — with new data points, new processes and usually new source systems. Starting the process design early is considerably cheaper than retrofitting it.

Don't you compete with Taxxor?

No — by design. Taxxor builds disclosure management software, and that is where I send anyone who needs tooling or filing production. My consultancy covers the surrounding questions: what your obligation actually means, how reporting connects to your systems and processes, and how your team prepares. Where an engagement would overlap with Taxxor's offering, I hand it to Taxxor.

Do I need a fractional architect or an agency?

An agency sells you a team to build something. A fractional architect sits on your side of the table: making the design decisions, reviewing what suppliers deliver, and keeping the landscape coherent across projects. If you already have builders but decisions keep stalling or going wrong, you need the architect, not more builders.

What does a fractional engagement look like in practice?

Typically one or two fixed days a week, embedded in your team: architecture decisions, integration designs, vendor conversations, code and design reviews. You get senior judgment continuously without a full-time salary, and the engagement ends when the gap it fills is closed.

Is my data safe with AI tools?

It depends entirely on which tools and how you deploy them — that's the honest answer. Public cloud AI services process your prompts on someone else's infrastructure under their terms. For genuinely sensitive data there are local and on-premise options where nothing leaves your environment. Part of my AI advisory is mapping which of your data can go where, before any tool is chosen.

Do you build AI solutions or only advise?

The core deliverable is advisory: a process scan, a tool selection, an adoption plan your own team executes. I'm a developer by trade and work with this technology daily — cloud models as well as local ones (Ollama, LM Studio) — so I can guide implementation hands-on; recently I used AI tooling to map a legacy application and generate its migration plan. But I won't sell you a black box you depend on me to run.

What does an engagement cost?

There's no public price list, because the honest price depends on scope. The ladder above is how cost stays controllable: start with a single advisory session, scale to a project or a fractional arrangement only if the first step earns it. Project work is priced on the deliverable, not on hours.

Where do you work?

From Geldrop, the Netherlands — on-site in the Eindhoven / Noord-Brabant region, and remote for everything else. I work in Dutch and English (German is functional), and reporting-technology work in particular is location-independent.

Book an intro call

Book an intro call

Tell me what you're trying to ship — a reporting obligation, an integration, an AI plan — and I'll respond within 1–2 business days. The first conversation is free and without obligation.

LinkedIn

Johan Thijs

Location

Geldrop, the Netherlands — on-site in the Eindhoven region, remote elsewhere

Response time

Within 1–2 business days, with a concrete proposal for the intro call