Restaurant · St Julian's, Malta
Restaurant group, St Julian's
“Four venues, one backoffice, one Z-report a night. We were duct-taping spreadsheets before.”
The shape of the business
Four full-service venues across St Julian’s and Paceville. A shared central kitchen at one of the sites; a shared menu (with venue-specific specials); a single owner and two managers. Roughly 80 covers per venue at peak, full table service, KDS in each kitchen, Wolt + Bolt integration for the takeaway side.
This is the archetype where the old POS — a per-venue legacy system with no central reporting — was costing serious operator time. Reconciling four Z-reports a night, four times a week, into a single weekly P&L was eating the better part of two days a month between the owner and the bookkeeper.
What the workflow needed
The multi-site constraint pulls in features that a single-venue café never touches:
- One menu, four kitchens. A price change to the burger had to propagate to every venue without four manual edits. Specials per venue had to be additive, not destructive.
- One inventory, four venues. Stock at the central kitchen is the source-of-truth; transfers to each venue have to be recorded so the COGS per venue is correct.
- One backoffice, four sets of Z-readings. Every venue closes its own day, but the owner needs the aggregated daily total without four separate logins.
- RBAC that scales. A manager at venue 2 should see venue 2’s data, not venue 3’s. Cashiers see only their own venue. The owner sees everything.
- Aggregator orders routed correctly. Wolt sends one webhook; the order has to land at the right venue based on the address.
What the VIND-shape workflow looks like
- One tenant, four venues in the VIND data model. Menu items belong to the tenant; venue-specific pricing and availability are overlays.
- Central kitchen as a venue with stock-source flag. Daily transfers to each retail venue are first-class movements (not “edit the count”); inventory journals reconcile by month-end with no manual SQL.
- Backoffice dashboard with venue filter and aggregated daily-totals view. Z-readings list across all venues, with the Sunday-to-Saturday weekly P&L generated automatically.
- RBAC matrix enforced at both the gateway and each service. A manager’s JWT carries their venue assignment; queries are tenant- and venue-scoped at the data-access layer.
aggregator-bridgevenue routing based on the order’s delivery address against each venue’s polygon. The owner draws the polygons in the backoffice once; the routing follows.
The reconciliation that made the case
Before VIND-style consolidation, the bookkeeper would spend the first Tuesday of every month rebuilding the previous month’s P&L from four sets of Z-printouts and a stack of supplier invoices. Eight to ten hours, every month, of work that produced a number nobody trusted to two decimal places.
With the workflow above, the same P&L is a one-click export. The reconciliation went from 8–10 hours to 30 minutes, and the bookkeeper’s actual job — interpretation, anomaly-spotting, talking to the accountant — got the time back.
This is the workflow that motivates VIND’s multi-site backoffice surface and the reason RBAC is enforced twice in the stack — at the gateway and again at each service.
Honest caveats
This is a composite profile, drawn from conversations with multi-venue restaurant operators during 2026 H1. The “central kitchen with retail venues” model is common in the St Julian’s / Sliema corridor; the specific group structure described here represents the archetype rather than a single operator.
The four-venue group profile is also the most complex VIND workflow we ship out of the box. Larger groups are an Enterprise conversation — come talk to us and we’ll quote properly.
Want to be next?
We're looking for named launch customers. 15-minute demo, no card.
Talk to us