A deterministic collection queue for unpaid invoices (no scoring model)
PAID tool ($4.90 one-time), FREE DEMO with fictional data: https://gettidycents.com/product/invoice-recovery-desk An aging report tells you how old each unpaid invoice is. It does not tell you who to contact today. We b
PAID tool ($4.90 one-time), FREE DEMO with fictional data: https://gettidycents.com/product/invoice-recovery-desk
An aging report tells you how old each unpaid invoice is. It does not tell you who to contact today. We built that as a fixed, visible ordering instead of a score nobody can explain.
Status is derived, never typed (simplified sketch)
function status(inv, today) {
const open = inv.amount - inv.paid - inv.credited; // integer cents
if (open <= 0) return 'paid';
if (inv.disputedCents > 0) return 'disputed'; // separate queue
if (inv.promise && inv.promise.date >= today) return 'promised';
return today > inv.due ? 'overdue' : 'not-due';
}
A promise is shown as an expected payment, never as money received. Disputed amounts go to their own queue so they don't get the same reminder letter as a slow payer.
Today's queue
A simplified sketch of the idea (the real rule set is longer, and every position in the queue comes with its reason in words):
queue.sort((a, b) =>
b.daysOverdueBucket - a.daysOverdueBucket || // older first
b.openCents - a.openCents || // bigger first
a.lastContact.localeCompare(b.lastContact)); // longest silence first
Then the cash part: given a cash gap by a date, which unpaid invoices would close it if they were paid.
Demo
9 invoices, $13,576 open, 5 to contact today, each with the suggested letter (first reminder, repeat reminder, payment-promise check, instalment offer, dispute clarification).
The demo has fixed fictional figures; your own invoices (or a CSV/XLSX import) are in the paid version. No accounting software or bank connection needed. One of the tools for running your business at Tidy Cents β every result shows its working.
What's your rule for when a reminder becomes a phone call?
Originally published by Dev.to WebDev. Aggregated on AIWithGhost for educational purposes β full credit and traffic to the original publisher.