AI Agents for Multi Entity
Reconciliation
Entities rarely disagree about numbers. They disagree about definitions, and a mapped consolidation always balances anyway.
- Entities disagree about definitions rather than about figures, and a consolidation that maps mismatched definitions still balances.
- Capitalisation thresholds, provision policy and recognition timing are set locally and are rarely aligned, which makes group comparisons unreliable.
- Account mapping is built once and then decays silently, because nothing announces drift in a consolidation that always foots.
- An unmapped new local account falls into a default line, producing a growing group balance with no explanation behind it.
- Elimination failures cluster on transactions that are intercompany in substance and not flagged as such, which only a counterparty scan will find.
AI agents for multi entity reconciliation test whether entities are applying the same definition to the same mapped account, reconcile the mapping against the accounts actually in use, and scan third party ledgers for counterparties that are group entities. ChatFin checks the things a consolidation cannot check itself.
A consolidation always balances, which is the reason these problems persist. Mapping drift, definition differences and unflagged intercompany transactions all produce a group result that foots perfectly and cannot be explained at entity level. ChatFin tests the mapping and the definitions rather than only the totals.
Entities do not disagree about numbers, they disagree about meaning
Multi entity reconciliation is usually described as making the figures agree. In practice the figures are fine and the definitions are not, and a consolidation that maps mismatched definitions produces a group result nobody can explain.
| What differs by entity | Why it happens | Effect at group |
|---|---|---|
| Local chart of accounts | Statutory requirements differ | Mapping absorbs the difference |
| Revenue recognition timing | Local practice or local GAAP | Group revenue moves periods |
| Capitalisation thresholds | Set locally, rarely aligned | Cost or asset, inconsistently |
| Provision policy | Local judgement | Group provisions are not comparable |
| Cost centre structure | Grown organically | Group cost analysis is unreliable |
None of that is an error at entity level and all of it distorts the group view. ChatFin tests whether entities are applying the same definition to the same account, rather than only whether the mapped totals foot.
ChatFin tests whether entities apply the same definition to a mapped account, not just whether the mapped totals add up.
Mapping drift is silent and cumulative
The mapping from local accounts to group accounts is built once, usually carefully, and then decays. Nothing announces the decay because a mapped consolidation always balances.
ChatFin reconciles the mapping itself against the accounts actually in use each period and reports what is unmapped, newly mapped or mapped to something retired.
ChatFin reconciles the mapping against the accounts actually in use each period, which is the check that catches drift before it reaches the group result.
Elimination completeness is a question nobody asks
Consolidation systems eliminate what they are told to eliminate. Whether that is everything which should be eliminated is a separate question and it is almost never tested.
| Transaction | Should it eliminate | Does it usually |
|---|---|---|
| Intercompany sales flagged as such | Yes | Yes |
| Intercompany sales not flagged | Yes | No, invisible |
| Profit in closing inventory | Yes | Only if tracked |
| Intercompany dividends | Yes | Usually |
| Recharges booked as third party | Yes | No, miscoded |
| A sale to a newly acquired entity | From acquisition date | Often missed in year one |
The failures cluster in the same place: transactions that are intercompany in substance and are not flagged as such. ChatFin identifies counterparties in the third party ledgers that are actually group entities, which is the only way an unflagged intercompany transaction is found.
ChatFin scans third party ledgers for counterparties that are group entities, which is how an unflagged intercompany transaction is found at all.
Where this underperforms
Three.
Deeper in multi entity reporting
The use cases this hub covers.
Worked on this page
Multi entity reconciliation
Covered as sections here
Definition alignment · Mapping drift · Elimination completeness · Unflagged intercompany
Questions group accountants ask first
If the consolidation balances, what is actually wrong?
Balancing is not the test. A consolidation maps local accounts to group accounts and adds them up, so it will balance whether or not the entities mean the same thing by an account. Different capitalisation thresholds, provision policies and recognition timing all survive that process untouched. ChatFin tests the definitions rather than the totals.
How does account mapping go wrong over time?
Quietly. A new local account is created and never mapped, so it lands in a default line. An account is remapped locally and comparatives stop being comparable. An acquired entity gets a fast mapping that is never revisited. ChatFin reconciles the mapping against the accounts actually in use each period and reports what is unmapped or newly mapped.
How do you find intercompany transactions that were never flagged?
By looking in the third party ledgers. A recharge booked as an external cost, or a sale to a newly acquired entity in the first year, is intercompany in substance and carries no flag. ChatFin scans counterparties against a current entity register, which is the only way those are found.
Test the mapping, not the total.
ChatFin reconciles the account mapping against the accounts actually in use and reports what is unmapped, newly mapped or mapped to a retired group account.
It tests whether entities apply the same definition to a mapped account, and scans third party ledgers for group counterparties.
Bring your group mapping table and one period of entity ledgers.
Book a demoAll images are served from the Pexels content network. No third party product screenshots, vendor logos or licensed media are used on this page.
Part of this cluster
Bring your group mapping table.
We will reconcile it through ChatFin against the accounts your entities actually used this period and show you what is falling into a default line.