Rapier
Every line, tested.
How do you turn audit judgment into a repeatable test?
Work alongside auditors to translate testing procedures into SQL rules that can examine the full population.
Explore the walkthroughAudit teams have more transactions than they can examine manually. A useful analytics platform needs tests that auditors can explain and exceptions they can investigate.
I built Rapier in SQL alongside auditors, translating their testing procedures into repeatable rules. It ran across more than 900 engagements and 12 million journal lines.
- Audit engagements
- 900+Scale of platform use
- Journal lines tested
- 12M+Full-population journal testing
- Rules built with auditors
- SQLAuditor procedures translated into code
01 / The approach
The reasoning behind the rules.
Each test connects a business concern to a defined population, an explainable rule and evidence for auditor review.
- 01
Load the population
Bring the relevant ERP records into the testing dataset.
- 02
Define the rule
Work with auditors to express the test in SQL.
- 03
Run the test
Apply the rule across the relevant population.
- 04
Review exceptions
Inspect flagged records and use the evidence to support a conclusion.
02 / Explore in practice
Explore the audit rule library.
Select a rule to see the tested population, exception rate and records that need follow-up. A flag identifies work for the auditor to investigate.
The engagement and journal-line figures describe platform use. This interactive library uses fictional records and illustrative rule results, with no client source data.
Illustrative data · not client records. Rules authored in SQL with auditors — pick one to run the exception population.
Rule library / Choose an audit area, then a test
3 of 20 tests shown
AR ageing > 90 days
Open receivables past 90 days — credit and collectability risk.
43
4
Follow-up needed
The table shows matches from a curated fictional sample. The example ERP population for this rule is 8,420 records; that larger population is not loaded in this walkthrough. A population exception rate cannot be inferred from these sample flags.
| Invoice | Customer | Balance (USD) | Days open | Status |
|---|---|---|---|---|
| INV-88421 | ACME Retail | USD 142,800 | 128 | Open |
| INV-90102 | ABC Foods | USD 67,300 | 96 | Open |
| INV-82001 | ABC Wholesale | USD 98,400 | 214 | Open |
| INV-93301 | ACME Technology | USD 41,200 | 168 | Open |
Assessment and next steps
4 sample records match this rule. These are review candidates awaiting supporting evidence.
What the screen suggests
These open receivables are older than the 90-day screening threshold. Their age warrants a collectability review; the screen alone cannot determine whether a balance needs an allowance.
Evidence to obtain
Check subsequent receipts, customer correspondence, disputes, credit notes and the latest collection activity for each invoice.
What to do next
Ask the receivables owner to explain the outstanding balances and document collection plans. Assess any allowance using the supporting evidence and applicable accounting policy.
Conclude after follow-up: document what the evidence supports, whether an explanation clears the flag, and any unresolved items. These suggested steps illustrate audit reasoning; no follow-up investigation has been performed on the sample.