Skip to main content

Documentation Index

Fetch the complete documentation index at: https://help.daylit.com/llms.txt

Use this file to discover all available pages before exploring further.

The cash flow forecast gives you a forward-looking view of when money from open invoices is expected to arrive. Rather than relying on due dates alone — which tell you when payment should arrive — the forecast draws on AI-predicted payment dates to show when payment is likely to arrive, based on each customer’s historical behavior. This distinction becomes especially important for planning around customers who consistently pay late.

​
What the forecast shows

The cash flow forecast is built around a 13-week rolling window, starting from the next Monday after your request date. For each week in that window, the forecast displays three data series: Invoice inflows by due date (green). The total balance of open invoices whose due dates fall in that week. This represents the amount you are contractually owed during that period. Invoice inflows by predicted date (red). The total balance of open invoices that the AI predicts will actually be paid during that week, based on the invoice_pay_date_prediction insight for each invoice. For invoices without a model prediction, the due date is used as a fallback to ensure the forecast covers all open balances. Trailing 6-month average weekly inflow (black line). The historical average of weekly cash receipts over the past six months, calculated from your actual payment records. This baseline helps you judge whether a forecast week is above or below your typical collection rate. When the red (predicted) bars fall noticeably behind the green (due date) bars — or lag by one or more weeks — it indicates that a meaningful portion of your receivables are likely to be paid late relative to terms.

​
How AI payment predictions work

Daylit trains payment date prediction models on your invoice and payment history. For each open invoice, the model generates a predicted_date — the date by which it expects the invoice to be paid. These predictions factor in each customer’s past payment behavior: whether they typically pay early, on time, or late, and by how many days. Predictions are stored as insights attached to each invoice and are updated regularly as new payment data comes in. As more payment history accumulates for a given customer, the model’s predictions become more reliable.
Prediction accuracy improves over time. In the early weeks after connecting your accounting integration, the model may rely on due dates as a fallback for customers with limited payment history. As payments are recorded, the AI refines its estimates.

​
Finding the cash flow forecast

1

Navigate to Analytics

From the left navigation, select Analytics. The Analytics section contains cash flow and other financial summary views.
2

Open Cash Flow

Click Cash Flow within the Analytics section. The 13-week forecast chart loads automatically, starting from the next Monday after today.
3

Review the chart

The bar chart displays green and red bars side by side for each week, with the black trailing average line running across the chart. Below the chart, a customer breakdown table shows each customer’s predicted weekly inflows and their total predicted collections over the 13-week window.

​
Interpreting the forecast chart

Your customers are predicted to pay later than their due dates. The gap between the green and red bars represents collections that will likely shift into later weeks. Focus collection efforts on the accounts driving the largest gaps.
Your customers are predicted to pay close to their due dates, which indicates good payment discipline across your portfolio. The forecast weeks are likely to match your contractual expectations.
Predicted inflows for those weeks are below your historical average. This may reflect seasonality, a concentration of long-term-due invoices, or customers with poor payment history. Consider reviewing the customer breakdown table to identify which accounts are contributing.
The customer breakdown table below the chart shows each customer’s per-week predicted inflows. If one customer is dominating a specific week, that concentration is visible in the table rows. Concentration in a single week or customer creates cash flow risk if that customer pays late.

​
Customer breakdown table

Below the chart, a table lists every customer with open invoices that fall within the 13-week window. For each customer, you can see:
  • Their predicted cash inflow for each of the 13 weeks
  • Their total predicted inflow across the full 13-week period
The table is sorted by total predicted inflow, largest first. This makes it easy to identify which customers represent the biggest portion of your expected collections — and which ones to prioritize if you are managing cash flow closely.

​
Improving forecast accuracy

The forecast becomes more useful the more payment history Daylit has access to. A few things to keep in mind:
  • Customers who have never paid an invoice in Daylit will initially use due dates as predicted dates, since there is no behavioral data to draw from.
  • Customers with a long payment history — especially those who pay consistently early or consistently late — will have more accurate predictions.
  • Keeping your accounting integration synced regularly ensures the model has access to the latest payment records.
If you notice that predicted dates seem consistently off for a specific customer, check whether their recent invoices and payments are synced correctly in your accounting integration. Missing payment records can cause the model to underestimate how quickly that customer pays.

Dashboard overview

See all sections of the AR dashboard and how data flows between them.

AR aging report

Analyze overdue receivables by aging bucket and export aging reports.

Payment predictions

Understand how Daylit predicts payment dates for individual invoices.

AI insights

Learn how Daylit generates and updates AI insights across your portfolio.