That sounds like it should not need saying. It does, because a quiet three-day buffer is a common default and it produces the worst possible conversation: your figure and your customer's figure disagree about the same invoice, and neither of you can see why. Ours ships at zero days — and if your terms genuinely include a grace period, you set it and it is your number.
Alert, with the way to act on it
9 invoices past due · Open overdue list
Momentum is measured against this workspace's own recent trend, not against a target nobody set.
Revenue and collections are routinely reported as one topic and they are two. A business can be growing and running out of money at the same time, and the pairing below is arranged so you would notice.
Recent invoicing compared with this workspace's own preceding trend, expressed as a direction rather than a score out of a target. A target is a decision somebody in your business made and we do not know it; a change in your own rate of invoicing is a fact we can see.
Invoices past their due date, with a grace period that ships at zero days and is yours to set. This is the number that ends up in a conversation with a customer, so it has to be one you can defend line by line rather than one produced by an undisclosed rule.
What is owed in total, distinct from what is late. The two move independently and confusing them is how a business with excellent collections and a heavy order book convinces itself it has a problem, or the reverse.
How long money actually takes to arrive, measured from your invoices rather than from your terms. The gap between the two is usually the single most useful number on the card, because terms are what you agreed and this is what happens.
Late payment concentrated by customer, so the pattern separates from the incident. One customer thirty days late twice is a credit conversation; twenty customers three days late is a process problem on your side, and the two need completely different responses.
Who your revenue actually comes from, which is frequently not who your team spends its time on. Paired with the lateness signal it answers the uncomfortable version of the question: is our biggest customer also our slowest?
If your invoices are in more than one currency, the outstanding figure shows the largest single-currency total, labelled, with the others disclosed alongside rather than converted into it. That rule exists because this exact report once added dollars into shillings and printed the result with no currency on it. The full account of that, and what replaced it, is on Accounting controls.
This is the single most important thing to understand about Sales Insights, and it is the thing most “AI analytics” pages are careful not to say. Every figure on the page — every count, every total, every flagged record — is computed by an ordinary database query against your own data, with thresholds you control. A language model is used for exactly one job: turning that computed result into a paragraph a person can read quickly.
Counts, totals, ages and comparisons, run against your workspace under the same tenant scoping as every other screen. This step is deterministic: run it twice on unchanged data and you get the same answer twice. If the engine says nine invoices are overdue, nine rows matched.
What makes a request stale, a cost a spike, an entry large or a balance overdue is a number in your settings, not a judgement in a model. The defaults are listed below. Change one and the flags change with it, immediately and predictably.
Only then is a short plain-language summary generated from those facts. It is written on a background job so the page never waits for it, and it fails soft: if no provider is configured or the request fails, the summary is simply not there and every number on the page is unaffected.
The summary is stored with a fingerprint of the exact figures it was generated from. When those figures move, the fingerprint no longer matches and the summary is regenerated rather than served stale. This closes the failure that makes cached AI commentary dangerous: a confident paragraph describing last week, sitting directly above this week's numbers, with nothing on the screen indicating the two disagree.
Results are cached per workspace for a window you set — fifteen minutes by default — so the page is fast without being out of date, and a refresh action clears it when you want the current picture immediately.
These are the actual settings behind the flags on this page. They are per workspace, they are yours to change, and the defaults are chosen to be defensible rather than dramatic — an engine tuned to flag everything on day one is an engine nobody reads by day three.
| Setting | Ships as | What it decides |
|---|---|---|
| sales_overdue_grace_days | 0 | Days past the due date before an invoice counts as overdue. Zero by default so the figure matches the invoice. Set it to match your actual terms if those include a stated grace period. |
| cache_minutes | 15 | How long a computed result is reused before it is worked out again, with a refresh action that clears it on demand. |
| modules_enabled | sales: true | Whether sales contributes to insights at all, including the workspace-level narrative across modules. |
Every flag carries the way to act on it. A finding is not a number on a dashboard you then go hunting for — it is rendered with a label and a link to the screen where the work is done. An insight you cannot act on from where you are reading it is a report, and you already have reports.
This is the situation the card is arranged to make visible, because it is the one that reads as good news right up until payroll.
More invoicing than the preceding period. On its own this is the number that gets reported upward and the meeting moves on.
Eleven days of drift that nobody agreed to. Growth is being funded by the eleven days, which works until the growth stops.
Two customers account for most of the overdue value. That makes it a credit conversation with two people rather than a process change affecting everyone, which is a considerably cheaper answer.
The top-customers signal and the lateness signal name the same two organisations. Nobody wanted to notice that, and it is precisely the thing a card should say plainly rather than leave to be assembled from three reports.
Every one of those four is a query result. The model's only contribution would be to say it in a sentence, and if it were switched off you would still have all four numbers and the same conclusion.
What AWRA OpsHub does today
More we can add to your workspace
Where we point you to a specialist
The middle column sits on top of an engine that already computes, thresholds, caches and links. If collections workflow or a cash-collection forecast is the row that would change your month-end, that is a scoped conversation.
Cash flow split three ways, entries that are large for you, and anomalies.
Stockout risk on a horizon you set, dead stock, overstock and margin problems.
Late orders, stale requests, cost spikes and how concentrated your supply is.
All four modules on one card, and the workspace-level narrative across them.
The detailed view behind the overdue signal, by age bucket and owner.
One customer, everything: invoices, payments, tickets and history.
Help Center
Run inventory, procurement, assets, sales, and field work with approved AWRA guidance for setup, migration, integrations, security, pricing, and support.
