A sceptical senior reads a client's books looking for the entries that do not make sense. Fraud Detection runs that read across the whole voucher population, every time, whether or not something happened to catch anyone's eye.
40 forensic checks across the population
Audcrix runs 40 forensic checks over the synchronised ledgers and vouchers, looking for the patterns that deserve a second look:
- Negative cash balances — a cash book that went below zero on a day.
- Duplicate ledgers and duplicate payments — the same party twice, the same bill paid twice.
- Cash-limit breaches — transactions over the limits the law sets for cash.
- Benford's law anomalies — digit distributions that depart from what genuine figures produce.
- Round-number patterns — amounts that are suspiciously tidy for the kind of transaction.
- Back-dated and out-of-hours postings — entries made at times that do not fit the business.
Convergence, not a list of alarms
A single flag is often innocent. Several independent flags on the same party, ledger or period are not. The convergence engine brings the signals together, so the entries where different patterns meet rise to the top instead of being lost among hundreds of individual hits.
Honest about what it could not see
A forensic check is only as good as the data under it. Where a check needs data that has not been synchronised — vouchers for a period, a master that is incomplete — it says so with a gap flag. A silent check reads as a clean result, and a clean result on missing data is exactly the kind of assurance that should never be given.
Ready for SA 240
The output is built for the auditor's responsibilities under SA 240: each flag comes with the entries attached, so it can be followed up, explained or escalated without anyone re-running the analysis by hand. Journal-entry risk scoring in the Standards on Auditing module takes the same idea into the testing of manual journals.
Why run it on every engagement
Fraud rarely looks like fraud from a distance. Running the sweep on every client, over every voucher, means the unusual entry is surfaced by design rather than by luck.
Questions CAs ask
What does the fraud detection module look for?
40 forensic checks, including negative cash balances, duplicate ledgers and payments, cash-limit breaches, Benford's law anomalies, round-number patterns and back-dated or out-of-hours postings.
Won't it raise too many false alarms?
Individual flags are brought together by a convergence engine, so entries where several independent patterns meet are ranked first.
What happens when the data needed for a check is missing?
The check raises a gap flag saying what it could not see, rather than reporting a clean result on incomplete data.
Further reading
Related modules
- Accounting Checks — Completeness, cut-off, accrual and classification checks run over the whole Tally books — every verdict showing its basis, every finding exportable as a working paper.
- Detailed Scrutiny — Transaction-level voucher scrutiny across the whole Tally population — exception flags, period read from the narration, and prior-period items confirmed by the CA.
- Statutory Audit — Statutory audit software built around the Standards on Auditing — materiality, SA 530 sampling, SA 510 opening balances and confirmations, documented as the work is done.