Skip to main content
The Command Center unifies all customer communications across voice, SMS, web chat, and social media into a single intelligent inbox with AI-powered conversation analysis and automatic resolution.

​
Overview

Every customer interaction—regardless of channel—is captured, analyzed, and either resolved by AI or intelligently routed to your team with full context.

AI Resolution

78% of conversations resolved automatically without human intervention

Response Time

From 4.2 hours average to 8 seconds with 24/7 AI coverage

​
Supported Channels

The Command Center monitors and responds across four communication channels:

​
Voice (Phone Calls)

AI capabilities:
  • Answers incoming calls 24/7, including after-hours
  • Books, reschedules, and confirms appointments
  • Answers common questions about services and pricing
  • Escalates complex or sensitive issues to staff
Transcription: Every call is transcribed in real-time with speaker diarization (AI vs. Client vs. Staff). Example conversation: 
[AI] Thank you for calling Etienne Intelligence. How can I help you today?
[Client] Hi, I want to schedule a Botox appointment at your SoHo location.
[AI] I have availability next Tuesday at 10 AM, 2 PM, or 4 PM. Which works best?
[Client] 2 PM works perfectly.
[AI] Wonderful! You are confirmed for Botox at SoHo on Tuesday at 2 PM with Dr. Chen.
Metrics tracked:
  • Call duration
  • Sentiment (positive/neutral/negative)
  • After-hours flag
  • Revenue impact
  • Resolution status

​
SMS (Text Messaging)

AI capabilities:
  • Two-way SMS conversations
  • Appointment confirmations and reminders
  • No-show prevention outreach
  • Rebooking campaigns
  • Post-visit feedback collection
Example flow: 
[AI] Hi Emma, friendly reminder — your Hydrafacial is in 2 hours 
      at Williamsburg. Still coming?
      
[Client] Yes running a little late though, maybe 10 min

[AI] No problem! We will have everything ready. See you soon!
Response time: Average 8-12 seconds for AI responses

​
Web Chat

AI capabilities:
  • Embedded chat widget on your website
  • Instant answers to pricing and availability questions
  • Real-time booking without leaving the site
  • Lead capture for follow-up
Example interaction: 
[Client] Hi, is there a first-time client discount?

[AI] Welcome! New clients get 15% off their first treatment. 
     Our most popular intro service is the Hydrafacial at $250 
     (would be $212.50 with your discount). Want to book?
     
[Client] Yes at SoHo please!

[AI] Booked! Wednesday at 11 AM at SoHo Flagship. Welcome to Etienne!
Conversion rate: 65% of web chat sessions result in bookings

​
Social Media

Platforms supported:
  • Instagram DMs
  • Facebook Messenger
AI capabilities:
  • Responds to DMs within seconds
  • Handles booking requests
  • Answers pre/post-care questions
  • Manages negative feedback escalation
Example: 
[Client via Instagram] Saw your before/after pics! How do I book Botox?

[AI] Thanks for the love! I can book you right here. SoHo has 
     openings Thursday at 10 AM and 3 PM. Interested?
     
[Client] 10 AM!

[AI] Booked! Thursday 10 AM at SoHo for Botox ($450). DM me if 
     you need anything else!

​
Channel Performance Dashboard

Visualize conversation volume by channel with an interactive pie chart:
Channel distribution pie chart
Channel colors:
  • Voice - Purple
  • SMS - Purple
  • Web - Blue (primary)
  • Social - Orange
Insights:
  • Identify your highest-volume channels
  • Allocate AI resources effectively
  • Track channel mix changes over time

​
Resolution Breakdown

Understand how conversations are resolved:

​
AI Resolved (65-80%)

Conversations completed entirely by AI without staff intervention. Common scenarios:
  • Standard bookings
  • Appointment confirmations
  • Rescheduling requests
  • Pricing inquiries
  • Service information
Color: Blue (primary)

​
Escalated (15-25%)

Conversations requiring human expertise or judgment. Escalation triggers:
  • Medical concerns or adverse reactions
  • Billing disputes
  • Negative sentiment detection
  • VIP client flags
  • Complex multi-service requests
  • Requests outside normal hours/policies
Color: Orange (warning) Escalation process:
  1. AI detects escalation need
  2. Conversation routed to appropriate team member
  3. Full transcript and context provided
  4. AI summarizes issue for quick handoff

​
Abandoned (5-10%)

Conversations where the client stopped responding. Follow-up:
  • AI automatically attempts one callback/message
  • After 24 hours, conversation flagged for manual review
  • Client added to follow-up campaign
Color: Red (destructive)
High abandonment rates (>15%) may indicate AI response quality issues. Review transcripts to identify patterns.

​
Key Metrics

​
Total Conversations

Count of all customer interactions across all channels in the selected time period. What counts as a conversation:
  • Phone call (any duration)
  • SMS thread (grouped by client)
  • Web chat session
  • Social media message thread
Typical volume: 200-400 conversations per location per month

​
Average Response Time

Calculation: 
Sum of (first response timestamp - inquiry timestamp) / total conversations
Before AI: 4.2 hours (next business day for after-hours) With AI: 8-45 seconds Reduction: 85-95% Impact on conversion:
  • Responses under 1 minute: 78% conversion rate
  • Responses 1-60 minutes: 52% conversion rate
  • Responses over 1 hour: 23% conversion rate

​
AI Resolution Rate

Percentage of conversations resolved without human intervention. Formula: 
(AI Resolved / Total Conversations) × 100
Benchmark: 65-80% depending on business complexity Trend to watch: This should increase over time as AI learns from escalated conversations.

​
Revenue at Risk

Total value of appointments and bookings associated with conversations. Includes:
  • After-hours inquiries that would have gone to voicemail
  • Quick responses that prevented client from calling competitors
  • No-show prevention saves
  • Rebooking captures
Example calculation:
  • 47 after-hours bookings × avg 420=420 = 19,740 revenue at risk
  • Without AI: Estimated 75% loss rate = $14,805 saved

​
Calls Answered vs. Missed

Weekly call performance bar chart
Bar chart showing weekly trends:
  • Green bars: Calls answered (AI or staff)
  • Red bars: Calls missed (went to voicemail)
Before AI implementation:
  • 40-60% of calls missed during business hours
  • 100% missed after hours
After AI:
  • Less than 5% missed (only during system maintenance)
  • 0% missed after hours

​
AI Agent Status

Three specialized agents power the Command Center:

​
Conversation Analyst

Type: Analysis
Status: ● Online
Tasks: 1,247 Responsibilities:
  • Transcribes and analyzes voice calls
  • Extracts intent and sentiment
  • Identifies booking opportunities
  • Suggests upsells based on conversation context

​
Response Monitor

Type: Monitoring
Status: ● Online
Tasks: 2,381 Responsibilities:
  • Ensures all inquiries receive timely responses
  • Tracks response times across channels
  • Flags delayed responses for staff attention
  • Manages SMS and web chat conversations

​
Escalation Tracker

Type: Tracking
Status: ● Online
Tasks: 312 Responsibilities:
  • Monitors escalated conversations
  • Routes to appropriate team members
  • Tracks time-to-resolution
  • Ensures no escalations fall through cracks
The “Tasks Handled” counter resets monthly. These are cumulative counts of all actions taken by each agent.

​
After-Hours Coverage

One of the most valuable Command Center features: Standard business hours: Mon-Sat, 9 AM - 6 PM
AI coverage: 24/7/365 After-hours performance:
  • Average inquiries captured: 40-60 per month per location
  • Conversion rate: 72% (booked appointments)
  • Average revenue per inquiry: $420
  • Monthly after-hours revenue: 16,800−16,800 - 25,200
Example after-hours scenarios: 
[Client - 10:45 PM] Can I book laser hair removal for this weekend?

[AI] Hi Nina! Saturday at 2 PM at Stamford is available. 
     Would you like to book?
     
[Client] Perfect!

[AI] All booked! Saturday 2 PM at Stamford for Laser Hair Removal. 
     See you then!
Highlight your 24/7 availability in marketing materials. It’s a significant competitive advantage.

​
Conversation Inbox

Access the full inbox by clicking “View Inbox” from the Command Center overview.
The inbox displays all conversations with: Status filters:
  • In Progress (active conversations)
  • Escalated (requiring attention)
  • AI Resolved (completed)
  • Abandoned (no response)
Priority indicators:
  • 🔴 Urgent (VIP clients, medical issues, complaints)
  • 🟠 Pending (standard escalations)
  • 🔵 AI Handling (no action needed)
Conversation card details:
  • Client name and phone
  • Channel icon (phone/SMS/web/social)
  • One-line summary
  • Timestamp
  • Revenue impact (if applicable)
Example: 
🔴 Maria Santos                    📞 Voice
Booked Botox appointment for next Tuesday at SoHo
$450 revenue • 2m ago • SoHo

​
Conversation Transcript View

Click any conversation to see:
  1. Full transcript with timestamps
  2. Speaker labels (AI, Client, Staff)
  3. Sentiment analysis (positive/neutral/negative)
  4. Action timeline (booking created, reminder sent, etc.)
  5. Revenue attribution
  6. Related appointments in Zenoti
Staff actions available:
  • Take over conversation from AI
  • Add internal notes
  • Flag for follow-up
  • Export transcript

​
Staff Conversation Queue View

When logged in as staff, the Command Center shows:

​
Active Conversations

Only conversations escalated to you or in-progress conversations you’ve joined. Typical staff view:
  • 3-8 active conversations at any time
  • Sorted by priority (urgent first)
  • Real-time notifications for new escalations

​
Quick Stats

  • Active Conversations count
  • Your average response time
  • AI Handling count (informational)

​
Best Practices

When AI escalates a conversation:
  • Reason is always provided in the summary
  • Full transcript gives complete context
  • Client is told “connecting you now” - they expect quick response
Target: under 15 minutes for standard escalations, under 5 minutes for urgent
Review AI-resolved conversations weekly to:
  • See how AI handles common objections
  • Learn effective booking language
  • Identify knowledge gaps (AI couldn’t answer)
  • Improve your own conversion techniques
If AI escalates unnecessarily or makes mistakes:
  • Click “Report Issue” in transcript
  • Describe what should have happened
  • AI model improves from this feedback
Monitor the “After-hours inquiries captured” metric.This is revenue that was previously lost completely. Share wins with your team to build confidence in the AI system.

Dashboard

High-level conversation metrics

Scheduling

See booked appointments from conversations

AI Agents

Deep dive into agent architecture

Build docs developers (and LLMs) love

Get started for freeTalk to us