Statutory ऑडिट ऑटोमेशन।

No — it removes the need to sample purely because manual review couldn't reach the whole ledger. CORAA reads every voucher, every journal, every ledger and tests each against a defined rule set. Where the auditor deliberately chooses to sample under SA 530, for instance a procedure that specifically calls for it, CORAA runs Monetary Unit Sampling to the ICAI formula, with a seeded, re-performable selection. Both are legitimate; the auditor decides per procedure.

Every clause attempts a data-driven draft from the books and the working papers already done. Where the books settle the question, the clause arrives concluded, Yes, No, or a reasoned Not Applicable by entity type. Where only an auditor can conclude, physical verification, fraud, disputed dues, deposits compliance, the draft arrives Pending with the governing question and the source data, never a guessed answer. Nothing enters Annexure A until the auditor confirms it.

Related-party transactions are tested against the RPT schedule and the Sec 177 / Sec 188 approval and disclosure requirements, with candidates surfaced from ledger names, common directors, PAN and address matches, and transaction patterns that resemble related-party dealing. Every match is a candidate for the auditor's judgement, never an auto-asserted relationship — a no-register or an incomplete list is itself flagged rather than silently accepted.

Director remuneration is tested against the Companies Act ceilings and the board/shareholder approval trail. Loans, guarantees and security given by the company are checked against the Sec 185 / Sec 186 compliance conditions, with the terms, overdues and approvals traced to the underlying schedule so the conclusion is sourced, not asserted.

The auditor is, entirely. CORAA tests the population, drafts the mechanical clauses and presents the evidence behind each one — it never signs, and it never concludes the judgement calls. Every clause and every finding is a draft awaiting the auditor's confirmation, exportable to the working-paper file and re-performable under SA 230.

Data is hosted in India, handled in line with the DPDPA, on ISO 27001-certified infrastructure. Client data is never used to train shared or third-party models.