Bevatel voiceAgent is an AI-powered virtual agent designed to handle voice interactions autonomously, helping businesses scale and streamline their call workflows.
Whether operating as the first point of contact or running fully automated call centres, VoiceAgents enable teams to offload repetitive tasks and ensure consistent, high-quality caller experiences, which can operate 24/7.


How to Access AI Agent:

Step-by-Step:

1. Log in to your call centre account

2. Navigate to AI Agent in the left menu

3. Select My Agents
   
   

Explanation:
This screen shows that there are no agents created yet. Click on “Create new agent” to begin.

4- Create New Agent
define the identity and core behavior of a new AI Voice Agent. It includes agent instructions, greeting messages, voice selection, and knowledge base linking.

Steps to Create a New AI Agent

1. Agent Name

• What to Do: Enter a clear, descriptive name (e.g., “Customer Service English Agent”).
  
  

• Why: Helps identify agents quickly when managing multiple agents.
  
  

2. Agent Instructions

What to Do: Write a role description or persona or any instructions you want the agent to follow

Tip: Write full instructions that help to improve agent quality and performance.

Example:
You are an arabic helpful customer care agent for a telecom company that assists with billing, data plans, and complaints.
Please answer all questions using the attached document in the knowledge base

Why: This sets the context for how the AI should behave and what tone to use when responding.

Explanation:

Tone & Role: This clearly sets a helpful, customer-focused tone in Arabic.

Scope of Support: Limits the AI’s domain to billing, data plans, and complaint handling, avoiding irrelevant answers.

Source of Truth: Instructs the AI to stick strictly to the attached knowledge base content, ensuring accurate and approved responses.

3. Greeting Message

• What to Do: Write the first message the AI agent will speak when answering a call.
  Example:
  
  
  Hello! Thank you for calling Bevatel. I'm your Smart assistant. How may I help you today?
  
  
  

• Why: This is your caller’s first impression , make it friendly, professional, and informative.
  
  
  

4. Agent Language

• What to Do: Choose the primary language the agent will use (e.g., English, Arabic).
  
  

• Auto Detect Option: If enabled, the AI will identify the caller's language automatically and respond accordingly.
  
  

5. Voice Selection

• What to Do: Select the voice the agent will use.
  You may be able to preview each voice by clicking the play button.
  
  

• Why: Voice selection affects user experience; choose one that matches your brand's tone.
  
  

6. Agent Knowledge Base

The knowledge base is where your AI agent learns what to say. It reads and uses the information you provide here to generate accurate, helpful, and consistent answers for customers.

This is a critical step in ensuring the AI responds with up-to-date product, billing, or service information.

• What to Do: Click “Select documents” to upload or link your internal knowledge base (FAQs, procedures, service descriptions, etc.).
  
  

• Why: The AI uses this to provide accurate and consistent responses to caller queries.
  
  
  
  1- Click the “Select documents” button under the Agent knowledge base section.
  
  

2- A pop-up titled “Add to agent knowledge base” will appear.

Choose “Upload file” (default tab).

Click the “Upload file” button or Add text
Title → Enter a short, descriptive title for the entry.
Example: Billing - Grace Period Policy

Content → Write the full text that should be used to answer related questions.

Example:

Customers have a 3-day grace period after the due date to pay their bill without service interruption.

3- Browse and select your knowledge document(s).

• Supported formats: .PDF, .DOCX, .DOC
  
  

• Maximum size: 20 MB per file
  
  

• Maximum total: 10 files per workspace
  
  

4- Once the file is uploaded, click “Add to knowledge base”.

You’ll now see your document listed under the agent’s knowledge base

Example :

If your agent’s persona is:

"You are an Arabic helpful customer care agent for a telecom company that assists with billing, data plans, and complaints..."

You should upload documents such as:

• Billing FAQs (in Arabic)
  
  

• Data plans & pricing breakdown
  
  

• Complaint handling procedures
  
  

• Service Terms & Conditions
  
  

These documents will serve as the source of truth for the AI.

Routing Options (Tip for users)

 On the right side of the screen, you’ll see a helpful info box:

You can route calls directly to this agent or make it part of your IVR flow.

Example use cases:

• Route after-hours calls to the AI Agent
  
  

• In IVR: “Press 1 to talk to the AI Agent, press 2 to talk to a live agent”
  
  

 Click View in IVR to configure this (optional).

Connect VoiceAgent via IVR Routing

IVR configuration screen

1. Destination selection with "AI Agent" option
   
   

2. Example of agent selection (e.g., "Sales Agent")
   
   

3. Tip panel: Route calls to AI agent

📝 Purpose

The IVR (Interactive Voice Response) system lets you define how incoming calls are handled, routed, and responded to, including routing calls directly to your AI VoiceAgent.

Using IVR, you can:

• Direct calls to specific agents or teams
  
  

• Offer callers options (e.g. "Press 1 for support")
  
  

• Automatically send after-hours calls to AI agents
  
  

 How to Route Calls to an AI Agent via IVR:

Step 1: Open IVR Settings

1. Go to Settings > > IVR
   
   

2. Click on an existing IVR, or create a new one
   
   

Step 3: Create an IVR Entry

In the IVR Entries section:

Digits = 1, Destination = AI Agents, Agent = Sales Agent)

Step 4: Add More IVR Options (Optional)

Click “+ Add Another Entry” to:

• Route 2 to a live agent
  
  

• Route 3 to voicemail
  
  

• Route 4 to another AI agent
  
  

Step 5: Save Your IVR Configuration

Click Submit to apply your IVR routing logic.

📌 Example Use Case

In your IVR:

• Press 1 to speak with our AI Sales Agent
  
  

• Press 2 to talk to a live representative
  
  

💬 Pro Tip 

 You can use IVR to:

• Send after-hours or weekend calls to the AI agent
  
  

 Best Practices

Once your AI agent is connected via IVR, your callers can reach it automatically or by selecting it through a menu, streamlining support and saving live agent time.

• Back: Return to the previous screen
  
  

• Next: Continue to Agent Configuration
  
  
  

1- End call :

This configuration allows you to define conditions under which the AI agent is allowed to end a call automatically ,improving efficiency and avoiding dead-air time.

How to Configure End Call Behavior:

1. Navigate to the Agent Configuration tab (after setting up Agent Details).
   
   

2. Under Actions, toggle “End call” to ON.
   
   

• This gives the agent permission to end the call under defined conditions.
  
  

3. A configuration panel appears on the right with a Condition dropdown.
   
   

Available End Call Conditions:

Steps to Set a Condition:

1. Click the Condition dropdown.
   
   

2. Select one of the following:
   
   

• Max duration
  
  

• No response for a duration
  
  
  

3. (If prompted) Enter the specific value (e.g., 5 minutes, or 20 seconds).
   
   

4. Click Confirm to apply the configuration.
   
   

Best Practice Tips:

Additional Notes:

• You can edit or disable this configuration later at any time from this screen.
  
  

• Use this feature alongside IVR routing rules for better session control.
  
  

• This is especially useful in after-hours or automated support flows.
  
  

Best Practices :

• Use specific names if you manage multiple agents across languages or departments.
  
  

• Keep greetings short and warm, under 30 seconds.
  
  

• Regularly update the knowledge base to improve AI accuracy.

• Run a test on each voice to verify clarity and tone.

Voices :

The Voices menu allows you to choose the AI agent’s spoken voice from a list of available synthetic voices. These voices simulate human speech and are used in all interactions between the AI and callers.

 How to Use the Voices Menu

1. Navigate to the AI agent section in the left sidebar.
   
   

2. Click on the “Voices” menu.
   
   

3. You will see a list of available voice profiles. Each card includes:
   
   

• Voice Code & Name (e.g., MA Mazen, SA Sana, SA Sarah, GE George)
  
  

• A play button ▶️ to preview the voice sample.
  
  

Voice Selection :

📌 Note: This screen is for previewing voices only. The final voice must be selected while configuring or editing a specific AI agent under the Agent Details tab.

Knowledge Base:



The Knowledge Base tab allows you to:

• View all knowledge files linked to your AI agents
  
  

• Upload, remove or replace  documents
  
  

• Track upload status
  
  

• Search and manage the AI’s source of answers
  
  

This is where your operations team ensures the AI always provides accurate, up-to-date information to customers.

 Uploading a New Document

1. Navigate to AI Agent > Knowledge Base from the sidebar.
   
   

2. Use the Upload file tab (selected by default). Or add text to add text directly in text box.
   
   

3. Click Upload file to browse your computer.
   
   

• Supported formats: .PDF, .DOCX, .DOC
  
  

• Max file size: 20 MB
  
  

• Limit: Up to 10 files per workspace
  
  

4. Once uploaded, it will appear in the My documents list below.
   
   

💬 Status “Completed” means the file is processed and ready for use by the AI.

 My Documents Table – Column Descriptions

Actions Available

In the Actions column, you can:

• 🗑️ Delete the document from the workspace
  
  

• ⬇️ Download the document
  
  

•   Replace the document
  
  

• You can also use the Search bar to quickly locate a specific document.
  
  

•  Refresh the list by clicking the Refresh link at the bottom after uploading new files.

Best Practices:

 Example Knowledge Base Files

Once your documents are uploaded and marked as “completed,” they’re ready to be used by agents configured in My Agents.

Created Agents :

The My Agents screen is the central dashboard for viewing, accessing, and managing all AI VoiceAgents created within your Bevatel workspace.

This is where your operations or admin teams can:

• Monitor active/inactive agents
  
  

• Edit agent configuration
  
  

• Duplicate or delete agents
  
  

• Track last update and usage
  
  

Each card on the "My Agents" screen represents an individual AI agent with summarized information and quick actions.

Actions You Can Perform

Tips for Organizing Agents

Use Case Examples

💬 Did You Know?

You can create multiple agents for different departments, brands, or even languages and manage them all in this one dashboard.

Credit History (Monitor Usage ):

The Credit History section provides visibility into:

• Consumption of allocated credits or billing units
  
  

• Date and time of credit usage
  
  

• Cost control and budgeting insights
  
  
  
  
  

This is especially important for finance, operations, and stakeholders who need to track ROI and optimise AI usage across call centers.

Credit History Page:

Pro Tip

When launching a new AI agent, monitor credit usage daily for the first week to catch misconfigurations (e.g., stuck loops, long holds, etc.).

 Final Checklist

Before launching your AI Agent:

• Agent name, instructions, and greeting configured
  
  

• Knowledge base files uploaded and validated
  
  

• Voice selected and tested
  
  

• End call conditions configured (optional)
  
  

• Credit usage monitored