7 proprietary AI modules in production — each at qualified maturity, not magic for every workflow.
Each module ships in the codebase at pilot ready or BETA maturity. This page mirrors dashboard honesty labels — deterministic first, human-in-the-loop, no guaranteed ROI claims.
Seven modules
What each AI module actually does
Engineering tracker 22/22 means code shipped — not every workflow is LIVE for every tenant. Maturity badges match dashboard labels.
AI Restaurant Brain
Pilot readyDeterministic — rules + your workspace data
Daily briefing from hub, KDS, and inventory signals — not an autonomous manager. Review before acting.
/dashboard/today
Digital Twin
BETASimulation — planning model, not live telemetry
Station load simulation for planning — not a real-time mirror of your physical kitchen layout.
/dashboard/analytics/digital-twin
Universal Menu Engine
BETACross-channel menu model — sync maturity follows each integration's BETA/LIVE status
Channel sync depends on partner credentials and Integration Health — not guaranteed bidirectional inventory.
/dashboard/menu/universal
Food Cost AI
Pilot readyDeterministic — rules + your workspace data
Recipe-based margins require accurate yields and receiving data — not tax or audit advice.
/dashboard/analytics/food-cost
AI Purchasing
BETAEOQ-style reorder math — human approves every purchase order
Recommendations only — buyer approves POs. Not autonomous procurement or vendor negotiation.
/dashboard/inventory/purchasing-ai
Kitchen Camera AI
BETAPreview default — synthetic until stream configured
Camera-ready platform with configurable detection modules — not food safety certification or live vision on all stations by default.
/dashboard/kitchen/cameras
Benchmark Network
BETAAnonymized cohort comparisons when sample size allows
Early deployments may show limited benchmarks — not third-party audited industry rankings.
/dashboard/analytics/benchmarks
Copilot surfaces
Optional LLM and co-pilot paths
Separate from the seven core moats — preview maturity with human approval on drafts.
AI Operations Copilot
PreviewOptional LLM — verify before acting
Deterministic insights by default; optional narrative when OPENAI_API_KEY is configured. Human approval on drafts.
AI Restaurant Co-Pilot
PreviewPurchasing, labor, and margin scans — approval draft before execute
Scans operational gaps and surfaces drafts — nothing runs until you approve and execute.
Cross-cutting honesty rules
- Engineering done ≠ sales LIVE — pilot GO/NO-GO still applies until staging proof.
- Deterministic first — rules, structured DB signals, and math — not opaque LLM for core ops.
- Optional OpenAI for copilot — Restaurant Brain briefing is not GPT-dependent on core path.
- Show SKIPPED — Integration Health and Today briefing surface honest staging gaps in demos.
- Qualified wording — pilot ready, BETA, preview, deterministic over AI-powered magic.
Sales-safe
Safe umbrella claims
Use these in decks and demos when module-specific evidence is visible.
- 7 proprietary AI modules in production (qualified per module maturity)
- AI-assisted operations hub — demo shows Today + at least one moat with honest labels
- Deterministic daily briefing on Today Command Center
- Human-in-the-loop purchasing recommendations with approval gate visible
Do not say
Forbidden umbrella claims
Run copy through verify-claims before publishing. Method labels below are for internal reference.
- untouchable
- fully autonomous restaurant ai
- guaranteed margin
- guaranteed savings
- live computer vision everywhere
- production-certified ai for all tenants
- magic agi
- toast iq parity
Method types: Deterministic — rules + your workspace data · Heuristic — threshold alerts from ops signals · Simulation — planning model, not live telemetry…