SERVICESACCOUNTANCYAI for accounting and bookkeeping firms
Chasing documents, reading invoices, proposing entries and keeping an eye on every client's books, over and over. We build AI and integrations in the packages you already use. The judgement stays with the accountant.
ISO 9001 & 27001
Documents are read and prepared. The bookkeeper approves and keeps the judgement. Explore AI employees for finance →
What is AI for accountants?
AI for accountants is software that takes over the preparatory work across many client administrations at accounting and bookkeeping firms: requesting and tracking documents from clients, reading invoices and receipts and proposing an entry, flagging each period which administrations need attention, completing the file for client acceptance and collecting payroll changes. It works in the packages the firm and its clients already use, such as Exact Online, Twinfield, AFAS, Loket and Nmbrs. Entries outside the agreed rules, the judgement on the figures and the advice stay with people. At Your Tech Club every action is logged and what we build stays with the firm.
Built for clients including Tata Steel, M&G Investments, Digistore24, Toppers in Concert, Bar-le-Duc and Loogman
Three places in an accounting or bookkeeping firm where time disappears before there is anything to review.
Documents arrive late and incomplete
Receipts in the app, invoices by email, a bank statement that's missing. Before anything can be posted, someone emails and calls clients, one administration at a time.
The same work, for every set of books
Reading, coding and checking works slightly differently for each client. It comes back every month and has little to do with professional expertise.
Anomalies show up at the close
A missing bank statement, an unusual VAT code or overdue items that keep growing. With many administrations, you often only see it when the deadline is close.
How do you process client documents faster?
Purchase invoices, receipts and bank statements arrive by email, through a portal or an app. An AI employee reads them as soon as they arrive: supplier, amounts, VAT and references. It compares them with earlier entries for the same supplier and proposes the entry, with ledger account, VAT code and cost centre where needed, plus a short explanation.
What falls within the rules you set per administration can go through. The rest goes on a list for the bookkeeper, with the anomaly attached. If a document is missing, it asks the client for it specifically, with a reminder if nothing comes. The accounting package remains where entries are posted.
- Invoices and receipts read, including photos
- Proposed entry with an explanation, set up per administration
- Missing documents requested from the client
How do you see sooner which client needs attention?
At a firm with many clients, the risk isn't in a single entry but in what slips between administrations. Each period, an AI employee runs the fixed checks you agree on: are all bank statements in, were unusual VAT codes used, are a client's overdue items rising sharply, what hasn't been delivered yet.
The result is a worklist for the bookkeeper, showing per administration what stands out and why. No automatic correction and no opinion on the figures. That way the close starts with the administrations that need it, and you can see where every client stands without asking around.
- Fixed checks each period, for every administration
- A worklist of what stands out and why
- Progress per client visible without asking around
Can AI prepare client acceptance and payroll changes?
Before a new client is taken on, the file has to be complete. An AI employee fetches data from the Chamber of Commerce register through the KVK APIs, runs UBO and sanctions checks through the sources you use and requests missing documents. The complete file goes to a person, with every match and every ambiguity at the top. Whether the client is accepted is the firm's decision.
For firms that run payroll for employers, the same pattern applies to payroll changes. New starters, leavers, changed hours and allowances come in by email and form. The AI employee checks whether they are complete, asks the employer for what's missing and puts them in Loket, Nmbrs or AFAS as a draft. The payroll administrator approves before payroll runs.
- Chamber of Commerce data and checks in one file
- Every match on a list for a person
- Changes per employer as a draft in the payroll package
Not a new package, but less gathering, retyping and checking around the packages you already have. The professional work stays with you.
Opportunity first, then technology
We first look at where work gets stuck between client and entry, and what that costs, across all your administrations. Only then do we choose the solution.
In your own packages
Exact Online, Twinfield, AFAS, Loket, Nmbrs or another package stays the source. We build the integrations and the AI employee around it.
The accountant decides
Entries outside the rules, the judgement on the figures, the advice and accepting clients stay with people.
Set up and logged per administration
VAT codes, ledgers and journals differ per client. We record those rules per administration, and every action can be traced.
Even if an integration is enough
If a standard integration or a better set-up of the package solves it, we say so and don't build AI.
What we build is yours
Integrations, instructions and the AI employee itself belong to the firm. We are model-agnostic, so you aren't tied to one provider.
- projects delivered
- returning clients
- from 127 reviews
- ISO27001 and 9001 certified
Three ways to bring AI into your organisation, side by side.
- Works in your own systems, such as ERP, CRM and emailStandalone AI tool: partly · AI freelancer: yes
- ISO 9001 and ISO 27001 certifiedStandalone AI tool: varies · AI freelancer: rarely
- Starts with the process, not the technologyStandalone AI tool: no · AI freelancer: varies
- Tells you when AI isn't the answerStandalone AI tool: no · AI freelancer: varies
- A person decides where needed, with loggingStandalone AI tool: partly · AI freelancer: varies
- Strategy, design, software and AI in one teamStandalone AI tool: no · AI freelancer: no
- Management and further development after go-liveStandalone AI tool: via subscription · AI freelancer: varies
- Dedicated contacts, office in AmsterdamStandalone AI tool: no · AI freelancer: varies
Four steps. The details come from the audit, because a bookkeeping firm with many small clients works differently from a firm doing compilation and audit work.
- 01
AI Opportunity Audit in five working days
We map the work from document to entry across your administrations and calculate what it costs now. You get the three opportunities worth the most and what we would not do.
- 02
The foundation
Access to the packages, agreements on client data, permissions and retention, and logging. This is also where we record what may run automatically per administration.
- 03
The first process
Often reading documents and proposing entries, or chasing documents, but the audit decides. At first, everything goes past a bookkeeper.
- 04
Expanding
Checks across administrations, client acceptance or payroll changes are added on the same foundation, without starting over.
After go-live it stays under supervision
- MonitoringWe track exceptions per administration and per period, and notice when a client suddenly delivers differently.
- AdjustingNew coding rules or a new client are tested on earlier entries first, before they go live.
- Access and loggingPermissions follow the roles in the firm and the clients each person serves. Logging stays on.
- ExpandingNew tasks are added as soon as they demonstrably add value.
No account manager. Your first conversation is with someone who does the work, often with one of the founders. After that you work with a dedicated team from our office in Amsterdam.
- Office on the Duivendrechtsekade in Amsterdam
- ISO 9001 and ISO 27001, certified by Kiwa
- 500 projects delivered
- 4.8 from 127 reviews
What can AI do for an accounting or bookkeeping firm?
AI takes over the preparatory work that comes back for every set of books: requesting and tracking documents from clients, reading invoices and receipts and proposing an entry, flagging each period which administrations stand out, completing the file for client acceptance and collecting payroll changes. The accounting package stays the source. Entries outside the agreed rules, the judgement on the figures and the advice stay with the bookkeeper and the accountant.
Does this work with Exact Online, Twinfield, AFAS and Loket?
Yes, through the APIs of those packages, and with other packages too: we can connect any system. VAT codes, ledger accounts and journals are set up per administration. We record that mapping per administration, so a proposal doesn't go wrong as soon as a client uses an account differently.
May AI post entries in a client's books on its own?
Only within rules you set in advance, per ledger account and per amount. Everything else is a proposal with a short explanation, and an employee approves it before the entry reaches the package. Invoices that don't match are never silently approved; they go on a list with the reason.
Does AI help with client acceptance and anti-money laundering checks?
With gathering and recording, yes: fetching data from the Dutch Chamber of Commerce register, running UBO and sanctions checks through the sources you use, requesting missing documents and completing the file. A match goes straight to a person. That does not make you compliant with the Wwft by itself: responsibility for the client due diligence and the decision to accept a client stays with the firm.
Can AI prepare payroll changes for the employers we serve?
Yes. An AI employee collects an employer's changes from email and forms, checks whether they are complete, asks for what's missing and puts them in Loket, Nmbrs or AFAS as a draft. The payroll administrator approves before payroll runs. In Nmbrs, access through the REST API is arranged per employer, not with one master login; we include that in the set-up.
Can AI answer questions from clients?
Recurring ones, yes, such as whether a document has arrived or which documents are still missing, based on what is in your systems. An AI employee sorts the shared mailbox and passes the rest to the right colleague with a summary. Tax or substantive advice stays human work. For an organisation in the staffing sector we built an assistant that only answers when it finds the answer in the documents, with the passage and the page. We keep to the same principle here.
What don't you do?
We don't replace the accounting or payroll package. We don't compile annual accounts, give an opinion on the figures or give tax advice. AI does not accept or reject clients. And if a standard integration or a better set-up of the package is enough, we say so and don't build AI.
Where do you start, and what does it cost?
With the AI Opportunity Audit. In five working days we map what the preparatory work across all your administrations costs now, which three opportunities are worth the most and what we would not do. The audit has a fixed price, known in advance. Implementation after that runs on a monthly fee that depends on scope and pace.
Related services
Get in touch with us directly. We'll get back to you as soon as we can
Your message goes straight to our team in Amsterdam.