Using Claude to automate OTA booking and collections reconciliation

Reconciliation is high-volume, rule-based and just messy enough that rigid scripts break. That's precisely where AI earns its place.

The situation

A Dubai-based short-term rental business took bookings and collections through multiple OTA (online travel agency) partners.

Reconciling those bookings and collections back to the books is exactly the kind of work that consumes a finance team: high volume, repetitive, and inconsistent enough between partners that a fixed script struggles.

What we did

We used Claude to automate the reconciliation of bookings and collections against the OTA partner data.

Once that was working, we extended the same approach to handle most of the company's internal reconciliations as well.

Why this, and not a script

Traditional automation handles reconciliation well right up until the data stops being uniform — a partner changes a format, a field is missing, a description doesn't match. Then it fails and a person picks up the pieces.

That tolerance for mess is the reason to reach for AI here specifically, rather than applying it to finance work where a deterministic rule would do the job better and cheaper.

Other things we've built.

Have a problem that looks like this?

Most finance problems are architecture problems wearing a different hat. Tell us what's breaking and we'll tell you what we'd do.

Talk to Us