Query outstanding invoices with ChatGPT
With Ledger Botje you can query your outstanding invoices from Exact Online directly via ChatGPT. Instead of logging into Exact Online and navigating through reports, you simply ask ChatGPT: "Show me all outstanding invoices" or "Which invoices are overdue?" The AI connects to your Exact Online administration in real time and returns a structured overview.
You can make your queries as specific as you like. Ask for invoices from a particular customer, filter by amount or date, or request a summary of the total outstanding balance. ChatGPT understands natural language, so you do not need to know any query syntax or API parameters.
Generate a debtor overview with AI
Beyond individual invoices, you can also generate complete debtor overviews. Ask ChatGPT questions such as "Give me a debtor ageing report" or "Which customers have the highest outstanding balances?" The AI analyses your Exact Online data and presents a clear summary, grouped by customer, age category or amount.
This is particularly useful for credit management. You can quickly identify which customers consistently pay late, what the average payment term is, and where the greatest financial risks lie. All without creating manual Excel exports or custom reports in Exact Online.
Tips for invoice queries via ChatGPT
To get the best results when querying invoices, keep these tips in mind:
- Be specific: "Show outstanding invoices over €1,000 from the last 3 months" yields better results than "Show invoices."
- Use follow-up questions: start broad and zoom in. First request all outstanding invoices, then filter by customer or period.
- Ask for summaries: "What is the total outstanding amount?" gives you a quick overview without listing every individual invoice.
- Combine data: "Show outstanding invoices and compare them with payment history" provides deeper insight.
Want to learn more about what is possible? Read how AI accounting works with Exact Online or browse the prompt library for more example queries. More reporting options? Discover 7 reports you can generate with AI.
From overview to action
Pulling a debtor overview is the easy half. The question that matters comes after: who do you call today, and about what?
That is exactly where follow-up questions pay off. After the overview, ask: "Which of these customers is structurally late, and for which is this the first time?" That distinction sets the tone of your call - with the first you talk about the process, with the second about this one invoice.
A second one that works: "Set the outstanding amounts against that customer's revenue this year." An outstanding 8,000 euro means something different for a customer buying 200,000 than for one buying 12,000.
What not to do
Do not carry amounts over unchecked. If you are going to call a customer or send a reminder based on an answer, verify the underlying invoices. Not because the figures from the administration are unreliable, but because a summary makes choices you want to make yourself - has a credit note been settled, is there a payment arrangement?
Do not send reminders on autopilot. The tone of a reminder is a relationship decision. Let the assistant draft it and read it.
Where to go next
To automate further, read how to have reminders written from your bookkeeping. Ready-made questions are in the prompt library.