Loading content…
Francois Roux
CEO & Co-founder
A single accounting file arrived at the accounting firm for a statutory audit, and inside it sat the records of not one company but two. The client had recently completed a merger by absorption, and the entries of both entities now shared the same dataset: in France, the FEC (Fichier des Écritures Comptables), the standardised file of all accounting entries a company must be able to produce.
The engagement covered the absorbing entity alone. The absorbed entity had been left without a statutory auditor since the transaction, so signing off the accounts before the two sets of records had been separated and examined in their own right would have exposed the firm to a clear professional liability risk.
The predecessor practitioner had done the groundwork, apportioning each entity's entries across separate journals. An allocation key therefore existed, but putting it to use by conventional means was laborious: sorting the data, building Excel pivot tables, and rebuilding a trial balance journal by journal. These are exactly the checks that time pressure tends first to push back, and then quietly to drop.
The ground had been laid before the file was ever opened. The client record already held the company's SIRET number (the unique identifier carried by every French business establishment), together with a description of the merger. Taxxa.ai drew on this context to establish the nature of the engagement.
The accountant then supplied the allocation key in plain language, pointing to the marker that identified the absorbed entity's entries. Three instructions were enough: Taxxa.ai rebuilt two separate trial balances, each with its own consistency checks. The entire exercise ran inside Taxxa.ai, and Excel was never opened: no manual sorting, no pivot tables, no journal-by-journal reconstruction.
This is where the real value surfaced. Unprompted, Taxxa.ai flagged entries that belonged to the absorbed entity but had been posted outside its dedicated journals, the one blind spot an allocation key can never catch, because a key returns only what it is told to look for. Taxxa.ai laid these entries out, explained the reasoning behind them, and proposed the adjustments to make.
For the chartered accountant, this was the moment the tool proved its worth:
The Taxxa solution identified, on its own initiative, entries lying outside the specified journals but manifestly belonging to the absorbed entity. It set out the underlying logic and formulated the corresponding adjustments. This level of understanding of the file's structure was entirely unexpected.
The payoff is not only the spreadsheet hours saved. It is the distance between a check that actually gets done and one that slips because the clock runs out. In statutory audit work, that distance has a name: professional liability risk.
Three things came out of this engagement. A check taken through to completion, where the calendar might otherwise have forced its postponement. A level of completeness no manual process could match, catching entries that had slipped outside the dedicated journals. And an audit scope that was both controlled and fully documented.
What lingers is not just how fast the work was done, but how well the tool read the file: applying an allocation key, then recognizing its limits without being asked. The lesson is one of method: the richer the context set up front in the client record, the sharper the analysis that comes out of it.
For any firm in the same position, namely a post-merger file with two sets of records sitting side by side and an absorbed entity left without an auditor of its own, the approach offers a way to lock down the audit scope without pouring hours into a manual reconstruction.
Disclosure: Names have been withheld to preserve client confidentiality. This article reflects a genuine engagement carried out by an accounting firm in Paris, France, with the Taxxa.ai solution.
