India Audit Lens
Run automated audit and data-hygiene checks
Use Data Hygiene for deterministic exception checks over imported records. Use Audit (Beta) to create a versioned automated-audit run across the agents enabled for your firm. Both surfaces produce review inputs—not…
By India Audit Lens product team · Updated 2026-07-17
Run the data-hygiene review
Confirm that the engagement's books have been imported and normalised. Open Examination → Data Hygiene. On Dashboard, read Total Checks, Passed, Warnings and Failures. Open a category card or category tab. The workspace runs 40 checks across party master, sales, purchase, general-ledger/journal, inventory and payments data. Select a check and review its issue count and affected rows. Click a row when a detail drawer is available—for example to see every record in a duplicate group. Use search on larger populations and verify the source fields, issue description and financial…
Exception-table columns are generated from the check and therefore vary. Typical fields include party, PAN/GSTIN, invoice/voucher reference, date, amount, narration, HSN/rate, payment reference and the reason the record was flagged.
Capture or escalate a data exception
Open the relevant failing check. To capture the full server-side population, click Capture all (N) to Findings. For a subset, select rows and click Capture N selected. Alternatively use Raise Issue or Request Documents from the action bar. Verify the success message and the updated N in Findings badge.
Run and review Audit (Beta)
Availability and output depend on server-side firm feature flags. If no agents are enabled, a run may complete without workpapers or findings.
Open Audit (Beta) in the engagement. Confirm the intended engagement and that the required imported data is complete. Click Run Audit once. Leave the run in progress; do not start a replacement simply because it takes time. When complete, note the new audit version and review Workpapers and Findings. Use Search, Risk, Domain and Data quality only to triage the list. Click a finding to review the affected population, supporting transactions, what to verify, why it matters and linked workpaper/review state. Open Company Health for a high-level orientation and follow its linked finding chips…
Do not compare counts across versions without also considering data snapshot, feature configuration and run time. Do not copy a health status or Doctor narrative into a signed conclusion without review and corroboration.
Expected result
The deterministic check population and the selected automated-audit version are reviewed; evidence supporting real exceptions is preserved; false positives and data gaps are documented; and only human-reviewed matters progress to Findings, issues, document requests or conclusions.