AI Agent RAG Knowledge Base
Comprehensive knowledge base optimized for AI agent retrieval-augmented generation (RAG).
Purpose
This document serves as the primary knowledge source for Olympus Cloud AI agents. It provides structured information for RAG-based retrieval, enabling agents like Maximus (Cockpit AI), Minerva (Sales AI), and others to answer queries accurately.
Platform Overview
What is Olympus Cloud?
Olympus Cloud is an Autonomous Restaurant Operating System (aROS) that combines:
- Point of Sale (POS) - Order entry, payments, table management
- Kitchen Display System (KDS) - Order routing, ticket management
- Workforce Management - Scheduling, time clock, team communication
- AI Operations - Voice ordering, intelligent insights, automation
- Edge Computing - Offline-capable on-premise servers (OlympusEdge)
Platform Products
| Product | Domain | Audience |
|---|---|---|
| Restaurant Revolution | restaurantrevolution.ai | Restaurant customers |
| Olympus Cloud | olympuscloud.ai | Enterprise/workforce |
| Creators Revolution | creatorsrevolution.ai | Content creators |
Key Differentiators
- Offline Resilience - Full operation without internet via OlympusEdge
- AI-First Design - Voice AI, intelligent recommendations, automation
- Radical Flexibility - Policy engine adapts to any business model
- Multi-Tenant Architecture - Enterprise-grade isolation and scaling
Shell Types
Restaurant Shells (restaurantrevolution.ai)
| Shell | Users | Primary Functions |
|---|---|---|
| Staff Shell | Servers, hosts, bartenders | Order entry, payments, tables |
| Customer Shell | Diners (mobile app) | Browse menu, order, pay, track |
| Kiosk Shell | Self-service customers | Self-order, payment |
| KDS Shell | Kitchen staff | View tickets, mark complete |
| Drive-Thru Shell | Drive-thru staff | Voice AI ordering, payments |
| Digital Signage | Menu boards | Display menus, promotions |
Platform Shells (olympuscloud.ai)
| Shell | Users | Primary Functions |
|---|---|---|
| Platform Portal | Business users | Dashboard, analytics, settings |
| Platform Admin | NebusAI admins | Tenant management, support |
| Cockpit | Operations team | Monitoring, alerts, AI ops |
| Workforce | All employees | Scheduling, time clock, chat |
Creator Shells (creatorsrevolution.ai)
| Shell | Users | Primary Functions |
|---|---|---|
| Creator Studio | Content creators | Build AI personas |
| Audience Shell | Fans | Interact with AI personas |
API Reference Summary
Authentication
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/auth/login | POST | User login |
/api/v1/auth/logout | POST | User logout |
/api/v1/auth/refresh | POST | Refresh token |
/api/v1/auth/mfa/verify | POST | MFA verification |
Orders
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/orders | GET | List orders |
/api/v1/orders | POST | Create order |
/api/v1/orders/{id} | GET | Get order details |
/api/v1/orders/{id}/items | POST | Add item to order |
/api/v1/orders/{id}/send | POST | Send to kitchen |
/api/v1/orders/{id}/close | POST | Close/complete order |
Payments
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/payments | POST | Process payment |
/api/v1/payments/{id}/refund | POST | Process refund |
/api/v1/payments/{id}/void | POST | Void payment |
/api/v1/payments/tips | POST | Add tip |
Menu
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/menus | GET | List menus |
/api/v1/menus/{id}/items | GET | Get menu items |
/api/v1/items/{id}/modifiers | GET | Get item modifiers |
/api/v1/items/{id}/availability | PUT | Update availability |
Tables
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/tables | GET | List tables |
/api/v1/tables/{id}/status | PUT | Update table status |
/api/v1/tables/{id}/assign | POST | Assign server |
/api/v1/reservations | POST | Create reservation |
Workforce
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/schedule/shifts | GET | List shifts |
/api/v1/timeclock/punch | POST | Clock in/out |
/api/v1/timeclock/breaks | POST | Start/end break |
/api/v1/team/messages | POST | Send team message |
Common Questions & Answers
Orders
Q: How do I create a new order?
A: Use POST /api/v1/orders with location_id and order_type (dine_in, takeout, delivery). Then add items with POST /api/v1/orders/{id}/items.
Q: How do I send an order to the kitchen?
A: Use POST /api/v1/orders/{id}/send. This creates KDS tickets for all kitchen items.
Q: Can I modify an order after sending to kitchen?
A: Yes, add items with POST /api/v1/orders/{id}/items. New items create additional tickets. To remove items, use the void endpoint.
Q: How does order routing work? A: Items are routed to appropriate KDS stations based on item category and prep station configuration. Each station receives only relevant items.
Payments
Q: What payment methods are supported? A: Credit/debit cards (Stripe), cash, gift cards, Apple Pay, Google Pay, and split payments combining multiple methods.
Q: How do I process a split payment? A: Create multiple payment records against the same order. Each payment can be a different method. Total must equal order total.
Q: How do I handle tips?
A: Tips can be added during payment (pre-auth) or after (tip adjust). Use POST /api/v1/payments/tips with payment_id and tip_amount.
Q: How do refunds work?
A: Use POST /api/v1/payments/{id}/refund with amount. Partial refunds are supported. Full refunds available up to 90 days.
Tables & Reservations
Q: How do I check table availability?
A: Use GET /api/v1/tables?status=available for current availability. For reservations, use GET /api/v1/reservations/availability?date={date}&party_size={size}.
Q: How does the waitlist work?
A: Customers are added via POST /api/v1/waitlist with party size and contact info. When table is ready, they receive SMS notification.
Q: Can tables be combined?
A: Yes, use POST /api/v1/tables/merge with table IDs. Combined tables function as single unit for ordering.
Workforce
Q: How do employees clock in?
A: Staff use PIN or biometric in Staff Shell, or call POST /api/v1/timeclock/punch with action=clock_in.
Q: How are schedules created? A: Managers create shifts in Platform Portal or via API. AI can auto-generate schedules based on forecasted demand.
Q: How do shift swaps work? A: Employee requests swap, colleague accepts, manager approves (if configured). All via Workforce Shell or API.
Offline Mode
Q: What works offline? A: Order entry, payments (queued), time clock, KDS, and reporting (local data only). Full sync occurs on reconnect.
Q: How long can a location operate offline? A: Indefinitely with OlympusEdge. Local database and processing handles all operations. Payments queue and process on reconnect.
Q: How do I check sync status?
A: Use GET /api/v1/sync/status/{location_id} or check Sync Status in Platform Portal dashboard.
Sub-Pages
- Indexing - How indexing works, embedding models, document chunking
- Querying - Query patterns, retrieval strategies, re-ranking
- Entities - Entity definitions, key concepts, relationships
- Maintenance - Index maintenance, updates, monitoring
Related Documentation
- ACP AI Router - AI model routing
- Minerva AI - Sales AI platform
- Vision AI - Computer vision
- Agent Contexts - Agent personas and prompts