Voice-Guided Onboarding
The AI-powered Voice-Guided Onboarding Experience transforms restaurant setup from hours of manual data entry to a 10-minute conversational experience.
Overview
The Vision
"Upload a photo of your menu, a PDF of your floor plan, and a CSV of your inventory. Talk to our AI Assistant. We build your entire operating system in 10 minutes."
Unlike traditional POS setup that requires expensive implementation specialists, Olympus uses Universal Data Ingestion with Voice-Guided Resolution to onboard any restaurant automatically.
Core Capabilities
| Capability | Description |
|---|---|
| Universal Ingestion | Images, PDFs, CSVs, natural language |
| Schema Alignment AI | Maps any data format to Olympus schema |
| Voice-Guided Setup | Conversational resolution of ambiguities |
| Multi-Modal Input | Voice, camera, drag-and-drop combined |
Getting Started
Prerequisites
- Active Olympus Cloud account
- Admin or Owner permissions
- Supported browser (Chrome, Safari, Edge) or mobile app
- Microphone permissions (for voice features)
Accessing the Onboarding Wizard
- New Accounts: Automatically launched after signup
- Existing Accounts: Settings > Onboarding > Re-run Setup Wizard
- Additional Locations: Locations > Add Location > Use AI Setup
Voice Interaction Layer
Activation Methods
| Method | How to Activate | Best For |
|---|---|---|
| Push to Talk | Hold microphone button | Noisy environments |
| Wake Word | Say "Hey Olympus" | Hands-free operation |
| Text Chat | Type in chat panel | When voice isn't possible |
Voice Commands
Setup Commands:
"Start menu setup"
"I want to import my inventory"
"Help me set up my floor plan"
"Add my employees"
Resolution Commands:
"Yes, that's correct"
"No, change it to..."
"Skip this for now"
"Go back"
"Show me what you've done"
Navigation Commands:
"What's next?"
"Show my progress"
"Undo the last change"
"Start over with [section]"
Speech Recognition
| Feature | Specification |
|---|---|
| Engine | Whisper v3 (Workers AI) |
| Languages | 6 languages (English, Spanish, French, German, Portuguese, Chinese) |
| Accuracy | Optimized for restaurant terminology |
| Latency | Under 500ms response time |
Text-to-Speech
| Feature | Specification |
|---|---|
| Engine | ElevenLabs / PlayHT |
| Voices | Multiple personalities available |
| Customization | Match your brand voice |
Multi-Modal Upload Widget
Supported File Types
| Type | Formats | Use Cases |
|---|---|---|
| Images | JPG, PNG, HEIC | Menu photos, floor plans |
| PDFs | Invoices, rosters, menus | |
| Spreadsheets | CSV, Excel (XLSX) | Inventory, employees, menu data |
| Data Files | JSON, XML, DBF | POS exports, legacy systems |
Upload Methods
Drag and Drop:
- Open the upload widget
- Drag files from your computer
- Drop into the highlighted zone
Click to Upload:
- Click "Choose Files" button
- Select files from file picker
- Confirm selection
Mobile Camera:
- Tap camera icon
- Take photo of menu/document
- AI processes immediately
Voice Upload:
"I want to upload my menu"
[Camera opens]
"Take a photo"
[Photo captured and processed]
Real-Time Preview
After upload, you see:
- Parsed items with confidence scores
- Highlighted uncertain items
- Suggested corrections
- Accept/Reject controls
┌─────────────────────────────────────────┐
│ Parsed Menu Items │
├─────────────────────────────────────────┤
│ ✅ Signature Burger - $14.99 (98%) │
│ ✅ Caesar Salad - $12.99 (95%) │
│ ⚠️ "Truffle Fries" - $8.99 (78%) │
│ └─ Is this a side or appetizer? │
│ ❌ [Unable to parse item] (32%) │
│ └─ Please describe this item │
└─────────────────────────────────────────┘
Exception Resolution UI
When the AI encounters ambiguities, it presents them for resolution:
Types of Exceptions
| Exception Type | Example | Resolution |
|---|---|---|
| Category Ambiguity | "Is 'Wings' an appetizer or entree?" | Select category |
| Price Uncertainty | "Is the price $12.99 or $21.99?" | Confirm correct price |
| Item Duplicate | "This looks like an existing item" | Merge or create new |
| Missing Data | "No price found for 'Daily Special'" | Enter price |
| Format Question | "Is 'Patron' a tequila brand or customer type?" | Clarify context |
Resolution Interface
┌─────────────────────────────────────────┐
│ 🤔 I need your help with something │
│ │
│ I found "Truffle Oil" on your invoice │
│ but it's not on your menu. │
│ │
│ What should I do? │
│ │
│ [Add to Inventory] [Add to Menu] │
│ [It's an ingredient] [Skip] │
│ │
│ Or tell me: ________________________ │
└─────────────────────────────────────────┘
Voice Resolution
Simply speak your choice:
- "Add it to inventory"
- "It's an ingredient for the truffle fries"
- "Skip this one"
- "Actually, that should be truffle oil for the kitchen"
Schema Alignment AI
How It Works
The LangGraph-powered Schema Alignment Agent:
- Receives unstructured/competitor data
- Analyzes field names and values
- Matches to Olympus schema using semantic search
- Transforms data with confidence scores
- Requests human confirmation for uncertain mappings
Supported Source Schemas
Tier 1 (Native Support):
- Toast POS
- Square POS
- Clover POS
Tier 2 (File Import):
- Lightspeed
- TouchBistro
- Revel
- SpotOn
Tier 3 (Legacy Parsing):
- Aloha (DBF/XML)
- MICROS (Fixed-Width)
- Shift4
Field Mapping Examples
| Source Field | Target Field | Transformation |
|---|---|---|
| "Item Name" | name | Direct copy |
| "Price (cents)" | price | Divide by 100 |
| "Stock Count" | quantity_on_hand | Direct copy |
| "Cost/Unit" | unit_cost | Direct copy |
| "Category" | category_id | Lookup/Create |
Onboarding Workflow
Phase 1: Business Setup (2 minutes)
AI: "Welcome! Let's set up your restaurant. What's your restaurant name?"
You: "The Golden Fork"
AI: "Great! The Golden Fork. What type of cuisine do you serve?"
You: "Italian-American, casual dining"
AI: "Perfect. How many locations do you have?"
You: "Just one for now, downtown Austin"
AI: "Got it! Let me create your account... Done! Now let's build your menu."
Phase 2: Menu Import (3 minutes)
AI: "Do you have a digital menu I can import, or should we start fresh?"
You: "I have a PDF of our menu"
AI: "Perfect! Upload it or take a photo with your camera."
[User uploads menu.pdf]
AI: "I found 47 items. Let me organize them... I have a few questions."
AI: "Is 'House Salad' a side dish or a starter?"
You: "It's a side"
AI: "Got it. And 'Tiramisu' - is that $8.99 or $9.99? The image wasn't clear."
You: "Nine ninety-nine"
AI: "Perfect! Your menu is ready. Want to review it?"
Phase 3: Inventory Setup (2 minutes)
AI: "Now let's set up your inventory. Do you have invoices I can scan?"
You: "Yes, I have a Sysco invoice from last week"
[User uploads invoice]
AI: "Found 34 items from Sysco. I'll add them as inventory with costs."
AI: "I see 'Chicken Breast 10lb' at $42.99. Should I link this to menu items?"
You: "Yes, link it to the chicken parmesan and grilled chicken salad"
AI: "Done! I've set up recipe costs automatically."
Phase 4: Staff Import (2 minutes)
AI: "Time for your team! Do you use any scheduling software?"
You: "We export from 7shifts"
[User uploads 7shifts CSV]
AI: "Found 12 employees with their roles and schedules. Importing now..."
AI: "I see 'John' is listed as 'FOH'. Should that be Server or Host?"
You: "He's a server"
AI: "Perfect. All staff imported with their schedules!"
Phase 5: Floor Plan (1 minute)
AI: "Last step - your floor plan. Take a photo of your dining room."
[User takes photo]
AI: "I can see about 18 tables and a bar area. Let me map them..."
AI: "I've created sections: Main Dining (12 tables), Bar (4 seats), Patio (6 tables)."
AI: "Does this look right?"
You: "The patio actually has 8 tables"
AI: "Updated! Your floor plan is ready."
Progress Tracking
Real-Time Progress Display
┌─────────────────────────────────────────┐
│ 🚀 Setting Up Your Restaurant │
├─────────────────────────────────────────┤
│ ✅ Business Information Complete │
│ ✅ Menu Import 47 items │
│ 🔄 Inventory Setup 78% │
│ └─ Processing invoices... │
│ ⏳ Staff Import Pending │
│ ⏳ Floor Plan Pending │
├─────────────────────────────────────────┤
│ Estimated time remaining: 3 minutes │
└─────────────────────────────────────────┘
Completion Summary
At the end of onboarding:
🎉 Your Restaurant is Ready!
Summary:
├── 47 menu items created
├── 15 categories organized
├── 34 inventory items tracked
├── 12 employees added
├── 18 tables mapped
└── 3 printers configured
What's Next?
→ Take your first order
→ Set up online ordering
→ Configure delivery
→ Customize your receipts
Competitor Migration
Quick Migration Path
For restaurants switching from other POS systems:
| From | Migration Time | Data Migrated |
|---|---|---|
| Toast | 5 minutes | Menu, Staff, Sales History |
| Square | 5 minutes | Catalog, Customers, Transactions |
| Clover | 5 minutes | Menu, Inventory, Employees |
| Aloha | 10 minutes | Full export processing |
| MICROS | 10 minutes | Legacy format parsing |
Migration Workflow
AI: "I see you're coming from Toast. Want me to import your data?"
You: "Yes please"
AI: "Download your Toast export from their portal, then upload it here."
[User uploads Toast export]
AI: "Perfect! I found:
- 89 menu items
- 2 years of sales history
- 24 employees
- 1,247 customer profiles
Ready to import everything?"
You: "Yes, import it all"
AI: "Importing... Done! Your entire Toast setup is now in Olympus."
Troubleshooting
Common Issues
| Issue | Solution |
|---|---|
| Voice not recognized | Check microphone permissions, reduce background noise |
| File not processing | Ensure file is under 100MB, try a clearer image |
| Wrong language | Say "Switch to English" or change in settings |
| Item parsed incorrectly | Use "Correct this" or tap to edit manually |
| Progress stuck | Say "Skip this step" and return later |
Getting Help
- Say "I need help" for contextual assistance
- Type "/help" in chat for command list
- Press F1 for keyboard shortcuts
- Contact support: support@olympuscloud.ai
Best Practices
For Fastest Setup
- Gather materials first - Have menus, invoices, schedules ready
- Use digital files - PDFs and CSVs work better than photos
- Good lighting - If using camera, ensure clear images
- Quiet environment - For best voice recognition
- Complete in one session - Progress saves, but momentum helps
For Best Results
- Confirm AI suggestions when prompted
- Provide context for ambiguous items
- Review the summary before finalizing
- Test with a sample order after setup
Related Documentation
- Data Ingestion API - Backend API reference
- Voice AI API - Voice processing details
- Schema Alignment - How the AI maps data
- Competitor Migration - Migration details