Voice Bot Features

Chat360 Voice Bot provides advanced capabilities for call analysis, conversation management, multilingual communication, and interaction intelligence. These features help businesses review voice interactions, manage conversation flow effectively, and gain actionable insights from customer calls.

By combining transcription, audio playback, sentiment analysis, and analytics, administrators can monitor bot performance and continuously improve the voice experience.

Call Review and Analytics

The call review feature enables administrators to analyze voice interactions using transcriptions, synchronized audio playback, and AI-powered summaries. This allows teams to understand conversations quickly while also reviewing key performance indicators for quality monitoring.

Key Capabilities

• Synchronized Call Transcriptions: View complete call transcripts aligned with audio playback for detailed analysis.

• AI-Generated Conversation Summaries: Automatically generate concise summaries highlighting the most important parts of a conversation.

• Sentiment Analysis: Analyze customer emotions to understand satisfaction levels and detect frustration or positive feedback.

• Call Metrics and KPIs: Display key metrics such as call duration, outcomes, and interaction details.

• Comprehensive Call Insights: Provides a holistic view of the interaction for better performance evaluation.

Example Use Case

A customer service manager reviews a support call by reading the transcript while listening to the synchronized audio. The AI-generated summary highlights key discussion points, while sentiment analysis helps identify customer satisfaction levels. Call duration and outcome metrics provide additional insights into agent performance.

Silence Handling Configuration

The Silence Handling feature manages bot behavior when callers remain silent during a conversation. It uses a configurable retry system to maintain the conversation flow and prevent calls from ending prematurely. This ensures a professional and smooth user experience even when users pause during the interaction.

Key Features

• Silence Threshold and Retry Delay: Configure the amount of time the system waits before detecting silence and triggering a retry prompt.

• Retry Attempt Management: Define the number of retry attempts the bot should make before ending the call. Setting the retry count to 0 disables retries.

• Retry Prompt Configuration: Create custom prompts such as:

  • "Are you still there?" or

  • "I'm here to help. Do you need more time?"

• Text-to-Speech Retry Prompts: Enable or disable retry prompts using TTS-based audio responses.

• Audio Preview: Preview the retry prompt audio to ensure correct playback before deployment.

• Professional Call Termination: After the configured retry attempts are exhausted, the bot can deliver a final message before ending the call.

Example Use Case

A healthcare appointment bot is configured with:

• 15-second silence threshold

• 3 retry attempts

• Custom prompt: "I'm here to help. Are you still looking for an appointment time?"

If the user remains silent, the bot retries three times before politely ending the call with a closing message.

Language and TTS Model Configuration

This feature allows administrators to configure multilingual voice bots with customizable text-to-speech settings and voice parameters. It ensures the bot delivers natural and culturally appropriate conversations across different languages.

Key Features

Multi-Language TTS Selection:

• Dropdown interface: "TTS Languages (1 selected)".

• Support for English, Hindi, and regional languages.

• Easy selection and management system.

TTS Engine Configuration:

• Choice between professional engines (Sarvam, Azure).

• Engine-specific optimization for different languages.

• Regional dialect support.

Auto Switch Language Intelligence:

• Toggle-enabled automatic language detection.

• Dynamic response switching based on customer's spoken language.

• Seamless conversation flow without interruption.

Advanced TTS Engine Options:

• Expandable configuration panel with gear icon.

• Deep customization settings for professional deployment.

• Comprehensive voice control options.

Custom TTS Prompt Configuration:

• Large text area for custom prompt entry.

• Personalized message creation for specific use cases.

• Flexible prompt customization.

Language-Specific Settings:

• Individual language configuration (English as default).

• Dedicated input fields for language customizations.

• Regional adaptation capabilities.

Voice Actor Selection:

• "Select Speaker" dropdown with multiple voice options.

• Brand-appropriate voice selection for different segments.

• Professional voice actor library.

Voice Parameter Controls:

• Pitch Control: Slider for tonal variation adjustment.

• Pace Control: Speaking speed customization.

• Loudness Control: Volume level fine-tuning.

• Precise slider controls for optimal customer experience.

Integrated Audio Preview:

• Built-in preview system (0:00 / 0:00 timeline).

• Play controls for real-time voice testing.

• "PREVIEW" button for comprehensive validation.

• Quality assurance before customer deployment.

Example Use Case

An international e-commerce support BOT serves customers across India and the Middle East:

• Configured with English and Hindi using Sarvam TTS engine.

• English: Professional female voice, moderate pitch/pace.

• Hindi: Warm male voice, slower pace for clarity.

• Auto-switch enabled for seamless language transitions.

When a Mumbai customer starts speaking Hindi after English greeting, the system detects the change and transitions to Hindi with pre-configured voice settings, maintaining context while adapting voice characteristics for localized experience.

Fallback and Interruption Handling

The Fallback and Interruption system allows the bot to handle customer interruptions naturally during conversations. Instead of ignoring user input while a prompt is playing, the bot evaluates the interruption and responds accordingly.

Configuration Purpose: Enable the BOT to:

• Handle customer interruptions during prompt delivery.

• Evaluate interruption relevance to current conversation context.

• Process valid responses and advance conversation flow.

• Re-prompt or redirect when interruptions don't match context.

• Maintain natural conversation rhythm based on live user input.

• Prevent rigid, scripted interactions that ignore customer intent.

Key Features

• Real-Time Interruption Detection: Detects when a user speaks while the bot is delivering a prompt.

• Contextual Evaluation: Analyzes whether the interruption matches the current conversation context.

• Dynamic Response Handling: Processes valid interruptions and updates the conversation flow accordingly.

• Intelligent Routing: If the interruption indicates a different intent, the bot can redirect the conversation to the appropriate flow.

• Natural Conversation Flow: Ensures conversations remain flexible rather than rigid or scripted.

Example Use Case

While the bot asks for an account number, a customer interrupts with:
"I want to cancel my subscription."
The bot recognizes the intent and immediately redirects the conversation to the subscription cancellation flow.

Keyword Analytics Configuration

Keyword Analytics enables administrators to monitor important words or phrases during voice conversations. This feature helps identify customer concerns, detect escalation signals, and analyze conversation patterns.

Configuration Purpose: Empower administrators to:

• Track specific business-relevant keywords and phrases.

• Monitor customer sentiment through keyword analysis.

• Identify common issues and escalation triggers.

• Configure threshold-based alerting for keyword frequency.

• Implement dynamic keyword detection using AI prompts.

• Generate analytics reports based on keyword occurrence.

• Improve customer service through conversation insights.

Example Use Case

A telecommunications company configures their BOT to track escalation-related keywords:

• Defined Keywords: "refund," "cancel," "supervisor" with synonyms like "money back," "terminate," "manager".

• Random Keywords Prompt: "Capture words related to escalation, like 'I want to talk to the manager,' 'issue is not getting resolved,' 'I want to escalate,' and any similar phrases".

• Threshold Setting: 2 mentions trigger alert.

When a customer says "This is terrible service, I want my money back and need to speak to your manager," the system captures "money back" (refund synonym) and "manager" (supervisor synonym), triggering an immediate escalation alert.

Feature Details

Define Keywords Section:

• Main Keyword Input: Primary keyword entry field (e.g., "Cars").

• Synonyms Management: Tag-based synonym system with removable tags (Mercedes, alto, ciaz, thar).

• Dynamic Synonym Addition: "Type and press Enter" functionality for easy synonym expansion.

• ADD Button: Simple addition of new keyword groups.

Random Keywords Section:

• AI-Driven Prompt Configuration: Large text area for natural language keyword detection prompts.

• Contextual Keyword Detection: Example prompt for escalation-related phrases.

• Dynamic Pattern Recognition: AI-powered identification of similar phrases and expressions.

• Flexible Keyword Capture: Ability to detect variations and contextual mentions.

Configure Threshold:

• Frequency-Based Alerting: Numeric threshold setting (shown as "2").

• Alert Trigger Management: System flags keywords when mentioned above set limit.

• Example Logic: "If limit is '3' and 'test drive' is spoken more than three times, system identifies and flags it".

• Customizable Sensitivity: Adjustable threshold for different keyword importance levels.

Allow Interruption Toggle:

• Interruption Management: Enable/disable customer interruption during keyword capture.

• Conversation Flow Control: Balance between keyword detection and natural conversation flow.

• User Experience Optimization: Prevent keyword tracking from disrupting customer interaction.

URL Training in Voice Bot

The Voice Bot supports URL-based training, which allows you to enhance the bot's knowledge by providing external web content directly — no manual data entry required.

Key Capability

• Add multiple URLs to train the bot

• No strict limitation on the number of URLs you can provide

• Continuously expand and update the bot's knowledge base at any time

How It Works

The bot fetches and processes content from the provided URLs and uses it to:

• Answer user queries accurately

• Improve contextual understanding

• Provide more relevant and precise responses during calls

Use Cases

URL training is ideal for pointing the bot to:

• Company websites — general business information

• Product documentation — features, specifications, and guides

• FAQs and help centers — common questions and answers

• Knowledge base articles — in-depth support content

Benefits

Scalable learning Add as many URLs as needed to keep growing the bot's knowledge without any restrictions.

Dynamic updates Keep the bot's knowledge aligned with your latest content — just retrain with updated URLs.

Reduced manual effort No need to manually input all data. Simply provide the URL and let the bot do the rest.

Important Notes

Ensure URLs are:

• Publicly accessible

• Contain structured and relevant content

Better quality content leads to better bot responses.

You can update or retrain URLs at any time to improve or refresh the bot's performance.

MCP Tool (API Integration)

Overview

The MCP (Model Context Protocol) Tool in AI Hub enables Voice Bots to connect with external systems through APIs. It allows seamless data exchange between the bot and third-party platforms such as CRMs, databases, or internal services.

Key Highlights

• Supports GET, POST, and PUT API methods

• Can be configured to run in two modes:

  • In-Call → Executes in real time during the conversation

  • Post-Call → Sends collected data after the call ends

• Fully customizable with:

  • Headers — Authorization, Content-Type, etc.

  • Query Parameters

  • JSON Body

• Enables use cases like:

  • Fetching customer data

  • Creating or updating CRM records

  • Validating user inputs in real time

Note: MCP tools make the Voice Bot dynamic and action-oriented — not just conversational.

Knowledge Base Training (URL Training)

The Knowledge Base Training feature in AI Hub allows you to train the Voice Bot using external content sources such as URLs, documents, or structured data.

Key Highlights

• Supports URL-based training for easy content ingestion

• Add multiple URLs without strict limits

• Helps the bot:

  • Understand business context

  • Answer FAQs accurately

  • Provide consistent and contextual responses

Benefits

Scalable learning Continuously expand bot knowledge by adding new URLs anytime.

Reduced manual effort No need to feed data line by line — just provide the URL and the bot handles the rest.

Up-to-date responses Keep training aligned with your latest content for accurate and relevant answers.

For further assistance, contact your platform administrator or refer to the AI Hub support documentation.