Tonight's service at a glance — every table, where it's up to, and the three things the floor needs to act on right now. The tablet surfaces; the floor decides. Run by Priya Sharma, Manager
Aisha Rahman (T7) has a tree-nut allergy on file. The smoked Tasmanian ocean trout is plated with fermented bunya nut, and the charred broccolini carries macadamia — both tree-nut. Flag the order before it reaches the pass; check with the guest and the kitchen.
Surfaced to Priya · taking T7's order nowThe lamington soufflé for T6 was fired 18 minutes ago and a soufflé won't hold. This is exactly the timing slip guests called out in a recent review — get Marcus to confirm the fire, or send word to the table.
Surfaced to Marcus Tan (sous) · echoes review follow-upDaniel Goh (T4) is in on his 6th visit — Manuka-smoked duck with the Bridge Road Robust Porter every time. A good table for Tommo to drop by; the floor decides whether and when.
Suggested for Tommo Finch · guest preference, uc06Regular status, the usual order and any allergy come straight off the guest profile — the same record the booking and CRM tooling keeps. That's how the tablet knew Daniel's porter and Aisha's tree-nut flag before the order was taken.
Guest profiles · UC06A pacing slip caught on the floor — like the T6 soufflé wait — gets logged against the service, so if a review later names it, the follow-up tool already has the context. Issues are surfaced live, then carried forward, not lost at the pass.
Review-response triage · UC04The floor stays in charge: the tablet surfaces preferences, allergies and pacing — it never speaks to a guest, fires an order or comps a table on its own. Every prompt is for a named person to act on or wave away.
This screen is a read-only floor view: it joins tonight's bookings to the guest record and shows table state, a simple pacing timer (time since seated, or since a course was fired), and any preference or allergy note already held against the guest. The pacing maths is plain arithmetic against typical turn and course times — there's no model speaking to guests and nothing fires automatically. The interesting work is making it sharper without ever taking the human out of the loop.