Vertical operating system · Built flagship

The operating layer running live for owner-led restaurants

Restaurant OS connects sales, labor, payroll, reviews, finance, and menus into one governed command surface — built end-to-end and live with SOI Thai Street Food & Sake Bar.

What's inside the pack

Restaurant OS ships with six operating modules. Sprint scope determines which modules go live for your operation — not every workflow is in every deployment.

Sales

POS activity into location-level summaries, channel mix, dish performance, and exceptions.

Labor

Timecards into labor summaries, overtime watch, missing punch queues, and payroll readiness.

Payroll

Weekly readiness into payroll runs, tip allocations, exceptions, and operator approvals.

Reviews

Reviews into insights and drafted responses, with approval routing and publishing.

Finance

Bank activity into a daily heartbeat, unclassified flags, and category routing.

Operator Brief

What changed across the operating layer, and what needs attention next.

Module 01 · Reviews Intelligence

Reviews become operating signals, not a dashboard

Reviews from every channel resolve into the same operating objects: topics, dish mentions, sentiment, and the exceptions worth a response. The operator sees what customers are actually saying about each location — not just the star average.

Objects: Review · ReviewInsight · ContentDraft · DishSalesSummary
Flagler location review intelligence — 4.7 rating, 117 reviews, top topics Food Quality, Service, Ambiance, and dishes mentioned including Pad Thai and Papaya Salad.

Location-level review intelligence · Flagler

AI Signals panel — durable patterns including 'Khao Soi remains Flagler-exclusive with 16 orders generating $314 revenue' and an AOV premium signal, each high-confidence and acknowledgeable.

AI Signals · durable patterns from the operations agent

Module 02 · Operator Brief

Durable patterns surface before the morning brief

The operations agent watches the operating layer continuously. Signals come with provenance, confidence, and source — and route into an acknowledge / dismiss workflow the owner can govern, not a feed they have to chase.

Objects: TrendSignal · OperatorBrief · Exception
Primitives used: Operator Brief, Approval Queue
Substrate · Beyond prebuilt modules

Compose new signals from business objects

Restaurant OS is not just six modules. It is a substrate of business objects — SalesTransaction, LaborSummary, CompensationPlan, Channel — and the operations and functions to compose new signals from them. Operators do not wait for a feature release to ask a new question.

Objects: Sales, labor, menu, channel, employee — the LWIS object model.
Operations: Filter · Join · Aggregate · Chart
Functions: Total Revenue · AOV · Effective Wage/hr · Tip %
Workflow canvas composing a 'Revenue by Location' signal from SalesTransaction through Filter, Aggregate, and Chart nodes.

Composing 'Revenue by Location' from the object model

Run your restaurant on the operating layer SOI runs on

A Restaurant Proof Sprint defines the workflow scope, the systems involved, and the control model before anything goes live. Sprint scope determines which modules ship for your operation — no two restaurants run the same shape.