home / use cases / back-office operations / payment-reminders

Payment reminders that send themselves

Every morning your Aivell checks the invoice ledger in your own database, finds what is overdue, and sends the right reminder — friendly at one week, firmer at three — in the customer's language. Payments recorded overnight stop the sequence automatically, and the final notice always waits for a human yes.

Watch the demo

starts when: Every weekday at 8:00, on schedule

The problem

Chasing money is the task everyone agrees matters and nobody wants to do. The invoice is booked, the due date passes, and the reminder waits — because writing “please pay” without sounding desperate or rude takes effort, and doing it thirty times a month takes an afternoon. Meanwhile the cash sits in someone else’s account.

So dunning happens in bursts. Once a quarter someone exports the aged-debtors list and sends a wave of reminders — a few of them to customers who paid last week, which is worse than sending nothing. It is textbook material for the 62% of employee time that goes to repetitive tasks: fully rule-based, endlessly postponed.

How your Aivell does it

  1. The ledger is read every morning. At 8:00 the chore queries your own invoice database — due dates, balances, payments recorded overnight. Whoever paid drops off the list before a word is written.
  2. The tone matches the delay. A week late earns a friendly nudge with the invoice attached; three weeks late gets firmer wording with accounting in copy. You define the ladder; the Aivell climbs it one rung at a time.
  3. Every reminder is personal. Written in the customer’s language, naming the actual invoice and amount — not a mail-merge blast from a noreply address.
  4. Your exceptions are respected. Key accounts stay off the email list, disputed invoices are skipped, and no customer hears from the box more than once a week, whatever the ledger says.
  5. The final notice waits for a human. Before anything with legal weight goes out, the chore stops. Elena gets the draft and the payment history on her phone — one tap sends it, one tap holds it.

The guardrails keep the whole thing honest. The database connection is read-only, so nothing the Aivell does can touch your books; email is physically limited to billing contacts already on file. The chore decides who gets reminded and when — the guardrails decide what is possible at all. Both are yours to edit, in plain language, whenever your collections policy changes.

The chore, as you'd write it.

No flowcharts, no code — a chore is just a message to your Aivell, in your own words. It shapes it into a solid, guarded procedure and follows it to the letter. This one:

Giulia Aivell · chase overdue invoices ● active

every weekday at 8, list invoices unpaid past their due date

drop anyone who paid overnight or has an open dispute

7 days late — friendly nudge, invoice attached

21 days late — firmer wording, cc accounting@

45 days late — draft the final notice and wait for Elena's yes

key accounts are never chased by email; put them on Elena's call list instead

write every reminder in the customer's own language

note each send in the dunning log so nobody is chased twice in a week

You
Got it — I'll start right away, and check with you whenever something needs a human.
Giulia Aivell

The tools it uses.

Connected with single-click passkeys — no copied tokens, no OAuth hell.

Database reads overdue invoices and overnight payments
Email sends reminders in the right tone
Schedule runs the round every weekday morning

Sensible guardrails.

Every tool can be limited to exactly what this chore needs. For this one, you might set:

Database Read-only on the invoice tables. It can look, never write — payments are booked by your accounting, not by the AI.
Email May write only to billing contacts already present in the database. A new or edited address needs a human first.
Email Sends from the reminders mailbox only; it cannot impersonate a colleague's personal address.

Questions, answered.

Will it pester customers who already paid? +

No — and that is the point of connecting it to your database instead of a mailing list. The chore re-reads the ledger every morning before writing a word, so a payment recorded overnight removes the customer from that day's round.

How does it know when to be polite and when to be firm? +

You wrote the ladder yourself, in the chore, in plain language — nudge at a week, cc accounting at three, final notice at forty-five days. Change the tone or the timing by editing a sentence, not a workflow diagram.

Is my customer and invoice data sent to a cloud service? +

Never. The box queries your database over your own network and does all its reading and writing onboard. Balances, names and payment history stay exactly where they already live — behind your firewall.

What if a customer replies to a reminder? +

Replies land in the reminders mailbox where your team reads them, and the chore can be told what to do — typically pause dunning for that customer and flag the thread, so nobody gets a firm letter mid-conversation.

Related.

What is an Aivell?

An Aivell is an on-premise AI colleague: a small box that plugs into your network, runs its own AI onboard and takes the repetitive work off your desk. Prompts, documents, data — nothing ever leaves your office. Unplug the internet: it keeps working.

You describe each task as a chore, in plain language. Your Aivell turns it into a solid, guarded procedure and runs it in the background — with guardrails on every tool, a complete audit trail, and approvals in your hand when you want them. One-time setup, fixed monthly fee. No tokens, no overages.

Watch the demo

onboard inference · air-gapped · fixed monthly fee