AI Assistant

The AI assistant processes customer messages across WhatsApp, Instagram, and Facebook Messenger, handling inquiries intelligently and automatically. Here's how the complete flow works:

1. Message Reception: Customer sends a message via their preferred platform (WhatsApp, Instagram, or Facebook Messenger). The message is received in real-time and queued for processing.
2. Context Analysis: The AI analyzes the message using natural language processing to understand intent, extract entities (dates, times, services), and maintain conversation context from previous messages in the thread.
3. Knowledge Base Retrieval: For questions or information requests, the AI uses RAG (Retrieval-Augmented Generation) to search your Q&A knowledge base, finding relevant information about services, pricing, policies, and procedures.
4. Tool Selection: Based on the customer's intent, the AI automatically selects and executes the appropriate tools: checking real-time availability, creating bookings, rescheduling appointments, cancelling bookings, or retrieving customer information.
5. Real-Time Data Access: The AI queries your live calendar, service database, staff schedules, and resource availability to provide accurate, up-to-date information.
6. Response Generation: The assistant generates a natural, conversational response that answers the customer's question or confirms the action taken, using information from your knowledge base and real-time data.
7. Booking Creation: If a booking is requested, the AI collects all necessary information (service, date, time, staff preference, customer details), validates availability, checks for conflicts, and creates the booking automatically.
8. Confirmation & Follow-up: Sends immediate confirmation messages with booking details. Automated reminders are sent before appointments based on your reminder settings.
9. Dashboard Logging: All conversations and actions are logged in your dashboard for review, allowing you to monitor AI performance and intervene when needed.

Advanced Features: The AI maintains conversation context across multiple messages, handles multi-turn conversations, asks clarifying questions when information is ambiguous, and can handle scenarios like multiple services and service combinations.

The AI assistant uses a Q&A knowledge base to provide accurate, context-aware answers. You add question-answer pairs in the dashboard; the system embeds them for semantic search (RAG). Here's how to build and maintain it:

• Q&A Pairs: Add question-answer pairs that cover what customers ask:
  • FAQs: Common questions about services, pricing, policies, business hours, location, and accessibility
  • Service info: What each service includes, duration, pricing, and what to expect
  • Policies: Cancellation, refunds, no-show, and other business policies
  • Business info: Hours, location, contact methods, and other details
• Embeddings: When you add or update Q&A pairs, they are embedded and indexed so the AI can find relevant answers quickly. You can trigger re-embedding when you edit a pair.
• Updates: Add, edit, or remove Q&A pairs anytime in the Q&A Pairs section of the dashboard. Changes are available to the AI after embedding.
• Contextual answers: The AI uses semantic search to find relevant Q&A pairs even when customers phrase questions differently, and combines that with conversation context and real-time booking data.
• Accuracy: The more relevant Q&A pairs you add, the better the AI answers. Update pairs when services, pricing, or policies change. Review conversation logs to find gaps and add new pairs as needed.

Configure your AI assistant in the dashboard. You can set the agent name, enable or disable tools, and activate or deactivate the agent. The AI uses your business profile, services, and knowledge base (Q&A pairs) to respond to customers:

• Agent settings: Name your agent and choose whether it is active. Each messaging channel (WhatsApp, Instagram, Facebook) can have its own agent.
• Tool Selection: Enable or disable specific tools for each agent:
  • Booking Tool: Enable automatic booking creation. Configure booking confirmation requirements, information collection preferences, and booking validation rules.
  • Availability Tool: Enable real-time availability checking. Configure how availability is presented to customers (time slots, date ranges, etc.).
  • Rescheduling Tool: Enable automatic rescheduling. Set rules for when rescheduling is allowed and how to handle conflicts.
  • Cancellation Tool: Enable automatic cancellation handling. Configure cancellation policies and refund processing.
  • Customer Information Tool: Enable customer data retrieval so the AI can access customer information when relevant.
• Testing: Use the Agent Playground in the dashboard to test your agent with sample questions. Verify that responses are accurate and that the knowledge base (Q&A pairs) is being used correctly. Test after adding or changing Q&A pairs.

Monitor and manage your AI assistant's performance to ensure it's providing excellent customer service and continuously improving. Regular monitoring helps you identify issues, optimize responses, and maintain high-quality customer interactions:

• Conversation History: View all customer conversations in your dashboard:
  • Access complete conversation logs for all customer interactions
  • View conversations across all platforms (WhatsApp, Instagram, Facebook) in one place
  • Search conversations by customer, date, topic, or keywords
  • Filter conversations by status, platform, or outcome
  • Export conversation logs for analysis or record-keeping
  • Review conversation flow to understand customer journey
  • Identify patterns in customer questions or issues
• Response Review: Review assistant responses for quality and accuracy:
  • Review individual responses to assess quality, accuracy, and helpfulness
  • Identify responses that could be improved
  • Flag incorrect or inappropriate responses for correction
  • Rate responses to train the AI on preferred response styles
  • Review responses that led to successful bookings vs. those that didn't
  • Identify common response patterns that work well or need improvement
  • Use response review to continuously improve AI performance
• Performance Metrics: Track response accuracy and customer satisfaction:
  • Response Accuracy: Track how often the AI provides correct, helpful answers. Measure accuracy by reviewing conversation outcomes and customer feedback.
  • Booking Conversion Rate: Track how many conversations result in successful bookings. Higher conversion rates indicate effective AI interactions.
  • Response Time: Monitor average response time. AI should respond within seconds for optimal customer experience.
  • Customer Satisfaction: Track customer satisfaction through feedback, ratings, or follow-up surveys. High satisfaction indicates effective AI service.
  • Question resolution: Review how often the AI successfully resolves customer questions from conversation logs.
  • Use metrics to identify areas for improvement and track progress over time
• Ban Management: Ban abusive conversations:
  • Identify and ban spam or abusive conversations from the Chats page
  • Block phone numbers or user IDs that send spam or inappropriate messages
  • Review banned users periodically and unban if appropriate
  • Protect your AI assistant from abuse while maintaining legitimate customer access
• Feedback Loop: Update knowledge base based on conversation patterns:
  • Review conversation logs to identify questions the AI couldn't answer well
  • Identify information gaps in your knowledge base
  • Add missing information to your Q&A pairs
  • Update existing Q&A pairs with more accurate or detailed information
  • Track which topics generate the most questions and ensure they're well-covered
  • Use conversation patterns to improve knowledge base organization
  • Regularly update knowledge base based on real customer interactions
• Monitoring: Review conversation logs to spot situations the AI struggled with. Update Q&A pairs and instructions to improve future responses.
• AI Training & Optimization: Continuously improve AI performance:
  • Review conversation logs weekly to identify improvement opportunities
  • Update knowledge base monthly based on new questions or information changes
  • Adjust AI configuration based on performance metrics and customer feedback
  • Test AI with new questions regularly to ensure knowledge base is comprehensive
  • Monitor industry trends and update knowledge base with relevant information
  • Get feedback from staff and customers about AI interactions
• Analytics & Reporting: Use analytics to understand AI performance:
  • View conversation volume trends over time
  • Analyze most common questions to identify knowledge base priorities
  • Track booking conversion rates by conversation type
  • Monitor response quality metrics and trends
  • Export analytics data for external analysis or reporting
  • Use analytics to make data-driven decisions about AI optimization

Manage customer conversations effectively to provide excellent service and maintain control over AI interactions:

• Real-Time Monitoring: View conversations as they happen in real-time. See customer messages and AI responses as they occur, allowing you to intervene when needed.
• Conversation Threading: All messages from the same customer are grouped into conversation threads. View complete conversation history in chronological order for context.
• Multi-Platform Conversations: See conversations from all platforms (WhatsApp, Instagram, Facebook) in one unified view. Switch between platforms seamlessly while maintaining conversation context.
• View conversations: View all conversations in the Chats page. For complex situations or complaints, you can manage the booking (reschedule, cancel, add notes) from the dashboard and contact the customer through your normal channels.
• Conversation Notes: Add internal notes to conversations visible only to your team. Record important information, follow-up actions, or special considerations.
• Conversation Search: Search conversations by customer, date, or keywords. Quickly find specific conversations or information.
• Conversation management: View, search, and filter conversations; ban or delete chats when needed.
• Conversation Review: View and search conversations to review how the AI handled interactions and identify areas for improvement.

The AI assistant includes advanced features for handling complex scenarios and providing sophisticated customer service:

• Multi-Turn Conversations: The AI maintains context across multiple messages in a conversation. It remembers previous messages, understands references to earlier parts of the conversation, and provides coherent, context-aware responses.
• Intent Recognition: Advanced natural language processing recognizes customer intent even when phrased differently. The AI understands variations in how customers express the same request.
• Entity Extraction: Automatically extracts important information from customer messages (dates, times, services, names, phone numbers) without requiring structured input.
• Clarifying Questions: When information is ambiguous or missing, the AI asks clarifying questions to gather necessary details before proceeding. This ensures accurate bookings and reduces errors.
• Booking scenarios: Handles multiple services in one appointment, service combinations, resource-specific bookings, and staff preferences with availability matching.
• Smart Suggestions: Proactively suggests alternative times when requested times are unavailable. Offers similar services when requested services aren't available. Suggests optimal booking times based on customer preferences.
• Contextual Understanding: Understands business context (business hours, service availability, staff schedules) to provide accurate, relevant responses. Adapts responses based on real-time availability and business rules.
• Error Handling: Gracefully handles errors, misunderstandings, and edge cases. Provides helpful error messages and suggests alternatives when requests can't be fulfilled.
• Learning & Adaptation: The AI learns from interactions to improve responses over time. Adapts to your business's specific terminology, processes, and customer communication patterns.

Best Practice: Regularly review how the AI handles complex scenarios. Update knowledge base and configuration based on edge cases. Test the AI with various customer scenarios to ensure it handles them effectively. Provide feedback through conversation review to help the AI improve.