A document portal changes nothing about a fiduciary firm's accounting competence; it changes the way documents arrive, get filed, and are found again at closing time. I have never built this portal for a client firm — this note is forward-looking, an architectural extension of FiscalDoc, the tool I built for my own tax affairs, applied to a profession that, by nature, handles the documents of dozens of third parties rather than its own. Here is what that change of scale would concretely involve, and what it would not solve.
The document-heavy daily life of a fiduciary firm, in four tasks
The fiduciary profession lives on documents that aren't its own. Four tasks recur, whatever the size of the firm.
Collecting client documents. Each client sends their supporting documents through whichever channel they prefer — email, post, drop-off at the office, sometimes a photo taken at the counter during a chance visit.
Follow-up. Chasing, once, then twice, then three times, the client who hasn't yet sent the expected document — often at the same time as everyone else, which concentrates the workload rather than spreading it across the year.
Filing. Sorting each document by client, type, and fiscal year, so it can be found six months or three years later, when someone needs it without warning.
Closing. Gathering everything, under the constraint of a tax or accounting deadline that doesn't move, regardless of the actual state of progress of the file when the deadline falls.
I am describing here a widely documented professional reality, not a mandate I would have carried out myself: I recognise it mainly because it resembles, on a reversed scale, what I experience with FiscalDoc — except that a firm lives it for dozens of clients at once, not for two personal scopes.
What would a document portal concretely change?
An owned portal would not change the nature of these four tasks; it would change their cost in time. Each client would deposit directly into their own space rather than through scattered emails — the document would arrive already attached to the right file. A classification engine, like the one running in FiscalDoc, would automatically identify the nature of the document and file it according to the firm's own logic rather than the generic logic of a rented tool. Follow-ups would trigger based on the actual state of a file — an identified missing document, not a forgotten one — rather than the memory of a staff member under pressure. And at closing time, the status of each client file would be readable at a glance, rather than reconstructed file by file, client after client.
What architecture? The multi-client extension of the FiscalDoc mechanics
Reduced to its essentials, a fiduciary document portal is a tool that would do, for several dozen clients, what FiscalDoc does for two scopes: collect, classify and retrieve documents, without any data passing through an uncontrolled third party. The architecture doesn't come from nowhere: it's the one I laid out in the anatomy of a custom-built solution, pushed one notch further. FiscalDoc manages two scopes for a single user, me. A fiduciary portal would manage one scope per client, for several staff members of the firm — which tips the project into the second trigger I described in that same note: multi-user capability. Each client should see only their own space; each staff member of the firm should see the whole, with a trace of who consulted or modified what. The interface would stay lightweight — a drop-off area, a dashboard — but authentication, role management, and the sealing-off between client files would become the core of the project, more so than automatic classification itself, which remains the easiest part to transpose from FiscalDoc. This is precisely the kind of project that gets scoped as a custom-built solution, not as a standardised product sold identically to every firm.
Why professional secrecy and the FADP weigh more heavily here than elsewhere
This is where the argument for local and owned stops being a matter of principle and becomes a central argument. A fiduciary firm doesn't handle its own data: it handles that of dozens of clients, often among the most sensitive there is — income, wealth, staff salaries, sometimes delicate financial situations. The profession operates under a confidentiality requirement specific to it, commonly referred to as the fiduciary's professional secrecy, formalised in the practices and codes of conduct of the sector. Added to this requirement is the FADP, which fully applies as soon as data belonging to natural persons passes through the firm's tools[1] — I detailed what the law concretely expects, including the register of processing activities and subcontracting, in Revised FADP: where does your SME's data really sleep?. A classification engine run locally or on controlled Swiss infrastructure meets both requirements at once: no client document passes through a third-party subcontractor, and the scope to document in a register of processing activities stays confined to the firm itself. This is exactly the principle FiscalDoc demonstrates at the scale of one person — the case is documented from end to end — applied here to a profession where confidentiality is not one option among others, but a condition of the mandate.
What a document portal doesn't solve
Honesty requires saying what this tool would not do. It would replace neither accounting judgement nor the advisory relationship: it would organise documents, not interpret them in a tax or strategic sense. Nor would it force adoption: some clients, particularly those less comfortable with digital tools, would keep sending their documents through the old channels, and the firm would have to live with both regimes in parallel for a long time. It would not eliminate follow-up: it would make it more precise, not unnecessary — some clients would ignore an automated reminder just as well as an email. And it would not be free to introduce, even though it would be free to operate once built: absorbing years of existing archives, training the team, running the old system alongside the new during the transition are real costs, of the same order as those I describe in the real cost of a SME's software subscriptions — except that here they would be incurred once, not every year.
What I describe here remains an architecture, not a turnkey offer: every fiduciary firm has its own flows, its own existing tools, its own tolerance for change, and a project of this type gets scoped before it gets built, like any custom-built solution. To place this project among a Swiss SME's other digital decisions, the overall roadmap remains laid out in What should a Swiss SME do about AI in 2026?
Key takeaways
— A document portal doesn't change the nature of a fiduciary firm's work; it changes the time cost of collection, filing and follow-up. — The architecture extends that of FiscalDoc, but tips into multi-user territory: one scope per client, authentication and roles to build. — Professional secrecy and the FADP weigh more heavily here than elsewhere: it's the central argument for a local or Swiss-hosted engine, under the firm's control.
FAQ
Does a document portal replace accounting software? No. It would organise the collection, filing and tracking of documents upstream; it would not keep the books nor produce the statements. The two tools would coexist, and their integration is precisely one of the projects that grows this kind of undertaking.
Would all of a fiduciary firm's clients agree to use a portal? No, and this must be planned for from the start. Part of the client base, often the oldest or least digitally equipped, would keep sending documents through existing channels. A realistic portal absorbs this heterogeneity rather than ignoring it.
Does professional secrecy forbid a fiduciary firm from using AI at all? No, but it strongly steers the choice of engine: a model run locally or on controlled Swiss infrastructure prevents a client document from passing through a third-party subcontractor. This is a trade-off to settle during scoping, not a renunciation of AI itself.
How much would such a portal cost? No general figure would be honest: the cost would depend on the number of clients, the level of integration with existing accounting tools, and the depth of follow-up automation. That's exactly what a scoping exercise must establish before any commitment.
And in your business? The AI Usage Diagnostic: sixty minutes to map your real flows, identify what deserves a custom-built solution, what stays in SaaS, and what needs no AI at all. Book a diagnostic
Sources
[1] Federal Act on Data Protection (FADP), revision of 25 September 2020, in force since 1 September 2023. www.fedlex.admin.ch/eli/cc/2022/491/fr [↩]
Jérôme Deshaie is CEO and founder of MCVA Consulting SA, an augmented agency based in Valais. Fifteen years serving major international brands, now working directly with Swiss SMEs. Background.