- General Metrics: A high-level overview of sessions, users, messages, and agent performance.
- Topic Tracking: Insights into trending topics, FAQs, and automated conversation themes based on user messages.
- Activity Report: Detailed logs of interactions, including timestamps, channels, and response times.
- NPS Report: Net Promoter Score analysis based on user feedback and satisfaction ratings.
General Metrics
The General Metrics section is a powerful command center designed to track the behavior of both AI and human agents. It allows you to discover trends, identify popular user topics, and make data-driven decisions to continuously improve your service. At the top of the dashboard, you can apply global filters to customize your view:- Date Range: Select periods of up to 365 days of historical data.
- Channels: Filter by specific platforms (e.g., Web, WhatsApp, Voice).
- Languages: View data for specific languages (e.g., Spanish, English, Portuguese).
- AI Agent: Select a specific bot to analyze its individual performance.
1. General (Main Dashboard)
This tab provides a comprehensive overview of message volume, user engagement, and satisfaction levels.Core KPIs
- Total Sessions & Unique Users: Track overall traffic and individual reach.
- Message Breakdown: Monitor volume for User Messages, AI Agent Messages, Human Agent Messages, and System Messages.
- Total Messages: The sum of all interactions in the period.
- Efficiency Metrics: Average messages per session and average session duration.
- Satisfaction (Ratings): Track user feedback with Positive, Negative, and No Rating counts (including percentages).
Conversation Context
- Conversation Start: Understand how chats begin—whether User-initiated or via Push notifications (agent-initiated).
- User Distribution: Multi-dimensional charts showing messages by Channel, Tools used, sessions by Language, and sessions by Source.
Time and Performance Analysis
- Session Distribution (24h & Daily): Identify peak hours and daily traffic trends.
- Hourly Message Volume: See exactly when the highest load of messages occurs.
- Human Agent Response Time Analysis: A specialized chart to monitor how long it takes for human agents to respond once a chat is transferred.
2. Agents (Agent Analytics)
Focused on the transition from AI to human support and the productivity of your staff.Performance KPIs
- Assignment Metrics: Track how many chats were Assigned to Humans, distinguishing between Automated and Manual assignments.
- Team Involvement: See the number of unique agents involved in sessions.
- Session Health: Monitor Closures by Agent Timeout or User Timeout to identify friction points.
- Response Speed: Tracks “Quick Responses” (replies sent in under 5 minutes).
Productivity & Availability
- Agent Workload: Detailed bars showing chats assigned vs. chats closed per individual agent.
- Connection & Time: Monitor total connection hours and average chat time per agent.
- Availability Status: A real-time timeline visualizing agent states: Online, Break, Ending Shift, and Training.
- Temporal Activity: Line charts showing agent activity levels by hour and by day.
3. Voice (Voice Analytics)
Specific metrics for voice-based AI interactions to evaluate call quality and automation efficiency.Voice Volume & Outcome
- Call Summary: Total conversations, highlighting those Resolved by AI (without transfer) vs. those transferred to humans.
- WhatsApp Notifications: Tracks push messages sent as follow-ups to voice calls.
- Outcome & Source: Visual distribution of how calls ended and where they originated.
Call Quality & Efficiency
- Deflection Rate: Percentage of calls handled entirely by the AI without human intervention.
- Duration Analysis: Average call duration and a breakdown by ranges (e.g., < 1 min, 1-2 min, etc.).
- Human Escalation Rate: Percentage of calls that required a human agent.
Recontact Metrics
- Loyalty & Friction: Track the Recontact Rate (daily and total period).
- Repeat Callers: See the number of unique callers and identify those who called 2 or more times, helping you spot unresolved issues.
4. Human Agent (Real-Time Monitoring)
This tab provides a “Live” view of your human support operations, essential for supervisors to manage current load.Live Operations
- Active Queue: Real-time count of users currently assigned to agents and users waiting in the queue.
- Current Workload: A percentage representing the total capacity used by the active team.
- Transfer Dynamics: The current rate of AI-to-Human transfers.
- Wait Times: Average time a user is currently waiting to be served.
Agent Status Table
A detailed list of all human agents including:- Status: Instant view of who is Online or Offline.
- Individual Workload: Current active chats vs. maximum capacity (e.g., 2/5).
- Language Skills: The specific languages each agent is configured to handle.
Main Dashboard
Main Dashboard KPIs
This section highlights the core Key Performance Indicators (KPIs) that summarize the agent’s performance and user engagement over a selected time range:- Total Sessions: Total number of user sessions initiated.
- Unique Users: Number of distinct users who interacted with the AI Agent.
- User Messages: Total messages sent by users.
- AI Agent Messages: Total messages generated by the AI Agent.
- Human Agent Messages: Messages sent by a human agent (if handover occurred).
- Total Messages: Combined number of all messages in both directions.
- Average Messages per Session: Average number of messages exchanged in a session.
- Average Session Duration: Mean length of user sessions, measured in minutes.
User Distribution
The User Distribution section provides insight into how users are interacting with the AI Agent across different platforms and languages. It includes two pie charts:- Messages by Channel: Displays the proportion of messages exchanged through each channel, such as web, voice, WhatsApp, etc.
- Sessions by Language: Shows the distribution of sessions by the language used in each conversation.
Time Distribution
The Time Distribution section presents a set of visualizations that help you understand when interactions with the AI Agent are taking place and how the agent performs over time. It includes the following charts:- Session Time Distribution (24h): Displays session volume by time of day. The Y-axis shows sessions grouped by hour, and the X-axis represents the hours in 24-hour format.
- Daily Session Distribution: Shows the number of sessions per day. The Y-axis displays the session count, and the X-axis shows the date in yyyy/mm/dd format.
- Hourly Message Volume: Highlights how many messages are exchanged each hour. The Y-axis shows the number of messages, and the X-axis represents the time of day in 24-hour format.
- Agent Response Time Analysis: Displays the AI Agent’s average response time (in seconds) throughout the day. The Y-axis indicates the response time, and the X-axis shows the hours in 24-hour format.
Agent Analytics
Agent Request Metrics
Displays high-level data on how sessions are escalated and handled by agents. Metrics include:- Assigned to Human
- Assigned Automatically
- Assigned Manually
- Agents Involved
- Agent Timeout Closures
- User Timeout Closures
- Average Session Time with Agent
- Quick Responses (under 5 minutes)
User Distribution
Highlights the distribution of escalations by:- Channels with Highest Escalations
- Languages with Highest Escalations.
Agent Detailed Metrics
Provides agent-level performance data, including:- Chats Assigned by Agent
- Average Chat Time per Agent (minutes)
- Chats Closed per Agent
- Total Connection Time per Agent
- Agent Availability (hours). Different agent statuses are represented with distinct colors: Online, Break, Ending Shift, and Training. This visualization helps you clearly identify each agent’s availability throughout the day.
Agent Time Distribution
Visualizes the distribution of agent activity over time:- Agent Time Distribution by Hour
- Agent Time Distribution by Day.
Voice Analytics
The Voice Analytics tab provides insights into the performance and quality of voice interactions handled by your AI Agent. You can filter the data by start date and end date to focus on specific timeframes. These metrics are essential to evaluate both user experience and system performance in voice-based interactions. This tab includes the following sections:Voice KPIs
Displays key performance indicators related to voice calls, including:- Total Calls
- Average Call Duration (min)
- Total Call Duration (hours)
- Successful Scored Calls
- Success Rate (%)
- High Stress Calls (%)
- Average Talk Duration (seconds)
Average Score Distribution
Visualizes the distribution of average scores assigned to calls, helping assess overall call quality.Call Trends
Shows how voice call volume and performance evolve over time, helping you identify patterns, spikes, or drops in activity.Real Time
Live Metrics
Displays real-time figures for:- Users Assigned to Agents
- Users in Queue
- Agent Workload
- Bot to Human Transfer Rate
- Average Wait Time (min)
Agent Status
Lists individual agents with the following information:- Name
- Status (e.g., Online, Offline)
- Current Workload
- Languages each agent is assigned to handle
Topic Tracking
The Topic Tracking section is an AI-powered engine that automatically categorizes user interactions to help you understand what your customers are talking about. Instead of manual sorting, the AI analyzes conversations and groups them into a clear hierarchy: TOPIC ➔ FAQ ➔ LINKED MESSAGE This section is organized into four specialized tabs designed to give you both high-level trends and granular control over conversation data.⚙️ How the AI Processing Works
The system operates in smart batches to ensure accuracy. Every time your AI Agent receives 50 new user messages, it triggers an automated processing cycle.- Analysis: The AI reads the messages based on the instructions provided in your Prompt Topic (found in Agent Settings).
- Categorization: It automatically generates or assigns relevant Topics and FAQs.
- Linking: Specific user messages are linked to these categories for your review.
⚠️ Crucial Note: If the Prompt Topic field in your Agent Settings is empty, the AI will not process messages or create any categories.
📊 1. Topics Analytics (Temas)
This tab allows you to visualize the “Big Picture” of user interests through three key perspectives:- Topic Distribution: A visual breakdown showing the proportion of mentions across your different categories.
- Top Topics: A ranking of the most discussed themes, helping you identify what matters most to your users.
- Topic Evolution: A trend chart that shows how interest in specific topics grows or shrinks over time, allowing you to spot seasonal trends or emerging issues.
❓ 2. FAQ Analytics (Preguntas Frecuentes)
Focuses specifically on common questions and automated answers. It provides:- FAQ Distribution: Shows how frequently different FAQs are triggered relative to each other.
- Top FAQs: Identifies the specific questions users ask most often.
- FAQ Evolution: Tracks the performance and frequency of automated answers over selected periods.
💬 3. Conversation by Topic (Conversación por Tema)
This is your manual management area to ensure the AI is learning correctly. It is divided into two parts:- Unclassified Messages: A repository of messages that the AI couldn’t automatically categorize (due to lack of context or new subjects). You can manually review these and assign them to the correct topic.
- Message Re-classification: You can browse messages already assigned to specific topics to verify the AI’s accuracy or move them if needed.
🛠️ 4. Topic Configuration (Configuración de Temas)
The control center for your categorization logic. In this tab, you can:- Manage Topics: Create new themes, edit existing ones, or delete those no longer needed.
- Define Hierarchy: Set up how FAQs are nested under broader Topics.
- Data Integrity: If you delete a topic that contains linked messages, the system will ask you to reassign those messages to a different category, ensuring no valuable user data is lost.
🔍 Conversation Insights & Safety Net
In addition to the main tabs, the system provides a Refresh/Reprocess capability. If you update your Topic Configuration or want the AI to take a second look at “Unclassified Messages,” you can trigger a manual re-analysis to improve the overall quality of your reports. This structured approach ensures that you transition from simply “receiving messages” to truly “understanding user intent” through data-driven insights.Activity Report
- Date Range: Select specific timeframes for analysis.
- AI Agent: Filter by a specific bot to see only the interactions handled by that agent (e.g., “Order Status”).
- AI Agent (Agente IA): Identifies the specific AI bot that managed the session. This is particularly useful for organizations with multiple specialized agents, as it clarifies which bot was responsible for each user request.
- User (Usuario): Shows the identification of the user, typically including their name and contact number (e.g., WhatsApp phone number).
- Human Escalation (Transferencia a humano): Clearly indicates whether the conversation remained with the AI or was successfully transferred to a human agent.
- Metadata: Displays technical details of the session, such as the Language used (e.g., ES) and the Communication Channel (e.g., VOICE, WHATSAPP).
- Created (Creado): The exact timestamp (date and time) when the session was initiated.
User Report
The User Report provides a centralized directory of all individuals who have interacted with your AI Agent. This section is essential for understanding your audience, managing lead information, and tracking long-term user engagement.
To refine your search and focus on specific user segments, you can apply the following filters:
- Date Range: View users based on their activity within a specific period.
- AI Agent: Filter users who interacted with a particular bot.
- Conversation Channel: Narrow down the list by the platform used (e.g., WhatsApp, Voice, Web).
User Data Breakdown
The report table displays the following information for each user:- Name (Nombre): The identified name of the user. If the name was not captured during the interaction, it may appear as a placeholder.
- Conversation Channel (canal de conversación): The primary platform through which the user communicates with the AI (e.g., WHATSAPP, VOICE).
- Email (Correo): The user’s email address, provided if they shared it during a conversation or if it was captured via system integration.
- Phone Number (Número de teléfono): The contact number associated with the user’s account or session.
- First Conversation (Primera conversación): The historical date of the user’s very first interaction with the platform, helping you identify new vs. returning customers.
- Last Conversation (Última conversación): The date of the most recent interaction, allowing you to track current engagement levels.