In risk and compliance, knowing what happened is already too late

Thirty years in and around financial reporting teaches you one uncomfortable thing about risk: by the time it shows up in a report, it has already happened. The fraudulent payment cleared. The control failed. The reconciliation broke. The finding is written. A risk report is, almost by definition, a record of things you can no longer prevent.

That is the ceiling of traditional business intelligence, and risk and compliance is where you feel it most. BI is very good at telling you what happened, and it stops there. In most parts of the business, a beat late is an inconvenience. In risk, a beat late is a loss, a restatement, or a finding you get to explain to the audit committee.

Decision Intelligence is the shift that closes that gap. It does not stop at what happened. It reads the same data, works out why it is happening, and tells you what to do next while you can still act. Know what, know why, know what next. In risk and compliance that is the whole game: it is the difference between catching the anomaly before the run clears and reading about it in the post-mortem.

So here are nine risk and compliance signals that sit in your data before they reach a report, why your reporting catches each one too late, and what changes when decision intelligence puts the why and the what next in front of you in time. I will run one company's control environment through it to keep it concrete: a monthly close of about six business days, several thousand journal entries a period, and an external audit once a year.

1. The payment that should never have cleared

Start with the one that keeps controllers awake. Sitting in this week's payment run is a small cluster of transactions, call it 480k, with all the hallmarks of fraud or error: a newly created vendor, bank details changed at the last minute, round-number amounts, an approver acting outside their usual pattern. Nothing that screams. Everything that should be checked.

The fraud report is a reconciliation you run after the payments have gone out. That is the whole problem. It tells you what cleared, and by then the money has left the building.

This is decision intelligence at its most literal. Fraud Detection reads the payment stream before the run clears (know what), tells you why each flagged item looks wrong (know why), and tells you which to pull for review (know what next), while pulling them is still possible.

2. The expenses nobody has time to check

Every finance team samples expenses because no team can check them all. So a steady trickle gets through: duplicate claims, out-of-policy spend, amounts split to sit under an approval limit. In our company it runs to about 120k a year, and it clusters in a handful of cost centers where the habit has taken hold.

The expense report shows you totals by category and cost center. It does not surface the outliers inside those totals, so the leakage hides in the average.

Expense Anomaly Detection reads every claim rather than a sample, flags the anomalous ones with the reason each was flagged, and points you at the cost centers driving the pattern, so the checking effort lands where it pays.

3. The intercompany break that blows up the close

If you have ever run a multi-entity close, you know this dread. Somewhere in the intercompany accounts is a 1.2M imbalance that nets to nothing at the group level right up until you try to consolidate, at which point it stops the close dead while everyone hunts for the missing side of the entry.

The reconciliation surfaces the break when you consolidate, which is the single worst moment to find it, with the close clock already running.

Intercompany Reconciliation Anomalies catches the imbalance mid-period, with the entities and the entries behind it, so it is a routine fix on a normal Tuesday rather than a fire drill on close day.

4. The number that is moving for the wrong reason

Here is the subtle one, and it is pure know-what-versus-know-why. A KPI moves, say gross margin drops two points or DSO jumps, and the dashboard shows you the move. What it cannot tell you is whether that is the business changing or a posting error, a miscoded batch, a control that slipped. Same movement on the chart, completely different response required.

The KPI report shows the what. It has no opinion on the why, so a data-quality problem and a genuine trend look identical until someone digs, usually weeks later.

Financial KPI Anomaly Detection separates the real business change from the anomaly and points at the likely cause, so you are not treating an error as a trend or a trend as an error.

5. The accruals that come back to bite at true-up

Accruals are estimates, and estimates drift. When they drift too far, the true-up in a later period swings the P&L and someone has to explain a variance that was really just a bad estimate two months ago. It is not a control failure, exactly, but it lands on the same desk.

You find out at true-up, a period or two after the estimate was set, which is far too late to have set it better.

Accrual Accuracy Forecasting flags the accruals most likely to swing before you book them, with the drivers, so the estimate is tighter going in and the true-up is quieter coming out.

6. The close that is about to slip

The monthly close is supposed to be six business days. This month it is quietly trending to eight, held up by two reconciliations that are running late and an intercompany entry that has not landed. Nobody has said so yet, because the slip only becomes obvious partway through the close, when the options to fix it are gone.

A close status report tells you where you are once you are in the close. It does not tell you, a week out, that the close is going to slip and why.

Financial Close Prediction forecasts the slip before the close starts and names the bottleneck, so you clear the blocker in advance instead of discovering it on day four.

7. Walking into the audit exposed

With six weeks to the external audit, readiness is sitting at about 72 percent: specific documentation missing, a couple of controls without evidence, a reconciliation that will not stand up to a sample. None of it fatal if you deal with it now. All of it painful if the auditors find it first.

The trouble is you usually find the gaps when the auditors do, because nothing was scoring your readiness in the weeks beforehand.

Audit Readiness Prediction scores readiness ahead of the audit and tells you which gaps to close first (know what, know why, know what next, applied to the one deadline you cannot move), so you walk in prepared rather than exposed.

8. The regulatory gap you did not know you had

Regulatory compliance gets checked periodically, at quarter end, at a review, when someone remembers. Between those checks, things drift: a threshold changes, a filing lapses, an entity slips out of line with a rule that tightened. The gap is invisible in the space between checks, which is most of the time.

A point-in-time compliance review tells you where you stood on the day. It says nothing about the drift that opens up the day after.

Regulatory Compliance Scoring scores compliance continuously and flags the gap and the specific rule at risk as it opens, not at the next scheduled review.

9. The risk building across the whole picture

Individually, each of the above is a line item. Together they are an enterprise risk posture, and that is the number the board wants: liquidity headroom, covenant compliance, concentration, exposure. The problem is it gets assembled once a quarter for the board pack, long after the components started moving.

By the time enterprise financial risk is drawn together for the board, it is a quarter old, and risk that is a quarter old is risk you have been carrying blind.

Financial Risk Scoring keeps that posture live and decomposed by driver, so the board sees a current picture and you see which driver to act on before it becomes the headline.

The thread running through all nine

Every one of these is the same story. The signal was already in your data, and the report reached you a lap too late to prevent the loss, the finding, or the scramble. In most functions that lap is annoying. In risk and compliance it is expensive, and sometimes it is the kind of expensive you have to disclose.

This is exactly what Decision Intelligence is for, and it is why it is more than a smarter dashboard. Business intelligence tells you what happened. Decision Intelligence adds the two questions that actually protect you: why is this happening, and what do I do about it, in time to act. Know what, know why, know what next. Applied to risk, it moves you from detective to preventive: from explaining the fraudulent payment after the run to holding it before, from finding the control gap at the audit to closing it six weeks out, from reporting enterprise risk a quarter late to steering it live.

eyko reads your ledger, your transactions, your reconciliations, and your controls on a beat, works out the why, and hands your risk and finance teams the what next while it still changes the outcome. If you want to see it joined up, fraud, controls, close, audit, and enterprise risk read together, that is the Risk & Compliance page.