QAI-voQAI-vo

Finance / Journal Entry Risk Analyzer

QAI Journal Entry Risk Analyzer

Scans every journal entry in the period — not a sample — for the patterns auditors look for: weekend and after-hours postings, round-dollar amounts, entries just under an approval limit, self-approved entries, and more. Entries are ranked by how many flags they trigger at once, because one flag is noise but several on the same entry is worth a look.

About this tool⌄

Scans every journal entry in the period — not a sample — for the patterns auditors look for: weekend and after-hours postings, round-dollar amounts, entries just under an approval limit, self-approved entries, and more. Entries are ranked by how many flags they trigger at once, because one flag is noise but several on the same entry is worth a look.

How it works

  1. Upload your full journal entry extract for the period
  2. The tool proves the population is complete before testing it — no gaps, no unbalanced entries
  3. About a dozen risk tests run automatically across every entry
  4. Get a risk-ranked exception list with the specific attributes behind each flag — never an accusation, just the facts

Who it's for: Internal and external auditors performing required JE testing, and controllers who want a full-population risk check instead of a small manual sample.

Upload input files

Upload CSV or Excel (.xlsx). Excel is converted to the table the engine expects — no need to paste JSON.

Journal entries / GL

Required · No file selected

One row per JE line: entry_id, line_no, effective_date, posted_timestamp, account, account_name, debit, credit, description, posted_by, approved_by, source, reversal_of.

Authorized posters/approvers (optional)

Optional · No file selected

Single column: user — without this, entries approved by their own poster (self-approved) can't be flagged.

Holiday dates (optional)

Optional · No file selected

Single column: date — lets the scan flag postings made on holidays, not just weekends/after-hours.

Run parameters

These map to the upstream engine flags. Edit them to match your files (defaults are demo values).

Service conversation

Scoped to this service. AI explains verified results; accounting compute stays in the analysis engine.

Example: “Run the analysis and explain what still needs attention.”