Here’s an overview of our VoiceAgent capabilities and how they can be configured to enhance and automate your communication workflows.
User Level: Admin
Introducing VoiceAgents at CloudTalk
CloudTalk 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 conversations, VoiceAgents enable teams to offload repetitive tasks and ensure consistent, high-quality caller experiences.
With VoiceAgents, you can automate a wide range of tasks, such as:
Reaching out to leads with personalized outbound calls
Handling onboarding calls or surveys without tying up human agents
Collecting structured data during calls to feed into CRMs or internal systems
Answering FAQs or intelligently routing callers to the right department or person
VoiceAgents are highly customizable, integrating with your existing workflows, tools, and data sources to deliver efficient, context-aware conversations at scale.
Note: VoiceAgent functionality is currently not available for all accounts. If you do not see the VoiceAgents tab in your CloudTalk Dashboard and would like to explore this feature, please contact our Support team or your Customer Success Manager for assistance.
Billing Information
VoiceAgent calls are billed on a per-minute basis. Depending on your billing configuration, charges are either deducted directly from your CloudTalk balance or reflected in your monthly invoice.
To manage usage effectively, you can define a maximum call duration per agent. Once this limit is reached, the call will automatically terminate to prevent additional charges.
Agent List Page
You can access the Agent List Page in the CloudTalk Dashboard > VoiceAgents > Agents. This page provides an overview of all VoiceAgents configured in your account and allows you to manage them.
If a VoiceAgents plan is active on your account, information about it will appear at the top of the page. This section shows your monthly usage of minutes and the plan renewal date. If you need to adjust your subscription, use the Change plan or Cancel plan buttons in the top-right corner.
💸 To learn more about available VoiceAgent plans, visit the VoiceAgent pricing page.
Below the plan overview, you will find a list of all VoiceAgents in your account. For each VoiceAgent, the list displays the following information:
Name: includes the VoiceAgent name along with an indicator of the call direction and the template used during creation
Status: Shows whether the VoiceAgent is active or incomplete (for VoiceAgents whose configuration has not yet been fully completed).
Skills: the skills configured for that VoiceAgent
# of calls (past 30 days): the total number of calls handled by the agent in the last 30 days
Avg call duration: the average duration of the agent’s calls
From this list, you can also edit or delete a VoiceAgent.
Deleting a VoiceAgent from your account does not delete the call records associated with it.
Creating VoiceAgent
To create a new VoiceAgent, click + New VoiceAgent in the top-right corner of the VoiceAgents list page. This opens the template gallery, where you choose the type of agent you want to build.
Choosing a template
The template gallery is organized into two sections:
Manual Setup gives you full control over the agent's configuration from scratch. Two options are available:
Receptionist (Inbound): a starting point for building a front-desk agent that answers calls, takes messages, and routes callers.
Custom: a blank agent with no predefined behavior. Choose this if you want to configure everything yourself for either inbound or outbound calls.
Ready-made templates are grouped by category (Reception, Support, and others). Each template creates a fully pre-configured agent with a goal prompt, greeting message, skills, and scenarios already set up. Pick one and you have a working agent you can test immediately.
Tip: Templates are a starting point, not a locked-in setup. After creation, you can edit every part of the agent, including the goal, greeting, skills, and scenarios.
What happens after you choose
Both paths walk you through four steps: Profile, Skills, Knowledge, and Scenarios & Guardrails. The difference is how much is already filled in for you.
If you pick a ready-made template, the first step asks only for a name, language, and voice. The call direction, goal prompt, greeting, skills, and scenarios are all pre-configured by the template. You can review and change any of it after creation.
If you pick Receptionist or Custom from Manual Setup, the first step includes the full set of configuration fields: name, call direction, language, voice, personalization, goal, and greeting. You configure everything yourself.
In both cases, the remaining three steps (Skills, Knowledge, Scenarios & Guardrails) follow the same structure. Templates pre-populate them; Manual Setup leaves them empty for you to build from scratch.
Profile Creation
After creation, every VoiceAgent has an edit page where you can view and change the agent's identity and behavior. The available fields are the same regardless of how the agent was created.
Agent Name | The internal name of the VoiceAgent. This name appears in Call Logs, Analytics, and Integrations (for example, in HubSpot as the agent name). Choose a name that clearly identifies the agent's purpose or target audience.
Examples:
|
Call Direction | You can configure your VoiceAgent to handle inbound or outbound calls.
|
Outbound Number | When configuring an outbound VoiceAgent, you must choose which number the agent should use when making calls.
Select a specific number or set the option to Automatic. If set to Automatic, the system will automatically select a number from your account that matches the destination country of the call. The system only supports exact country matches. If no matching number is available, the Failover Number will be used instead. |
Failover Outbound Number | The Failover Outbound Number acts as a backup number used when no suitable local number is available. |
Language | The language your VoiceAgent should communicate in. Choose the language that best matches your target audience and use case. |
Voice | Choose the voice your VoiceAgent will use during calls. You can preview available voice options directly in the dashboard. Some voices include additional descriptors such as gender, accent, or tone. |
Personalization | For inbound agents, you can insert contact fields (such as |
Goal | The Goal defines the primary objective of the VoiceAgent during conversations. In this field, you should describe:
You can also include guidance about tone, conversation structure, or important information the agent should gather.
Example: The goal of this VoiceAgent is to confirm customer orders. During the call, the agent should verify the order details, ask the customer whether they would like to proceed with the order, and record whether the order is confirmed or declined. |
Greeting Message | Insert the message you want your VoiceAgent to use when starting every conversation. This greeting sets the tone of the call and introduces the purpose of the interaction. If filled in, the agent will speak first when a call connects. If left empty, the agent will wait for the caller to initiate the conversation. For Inbound agents, this field is required.
Example: "Hello, this is CloudTalk calling to confirm your recent order. Do you have a moment to review the details with me?" |
When you create an agent from a template, the goal and greeting are pre-written with placeholder company details. Replace these with your own information before going live.
Skills configuration
In the Skills step, you define what actions your VoiceAgent can perform during calls. Skills determine how the agent handles tasks such as collecting information, answering questions from a knowledge base, transferring calls, or booking appointments.
Template-based agents come with certain skills already enabled. For example:
The AI Receptionist template enables Take a message and Extract information by default, with Transfer to Human and Appointment Booking available but turned off.
The Appointment Scheduler template enables Extract information and includes pre-built custom skills such as Booking and Confirmation and Objections, Questions, and Refusals.
Skills with a ⚠️ warning icon require additional setup before they will work (for example, Transfer to Human needs a transfer target, and Appointment Booking needs a connected calendar).
You can enable, disable, or edit any skill, and add custom skills using the + Add option at the bottom of the list.
For detailed guidance on each skill, see the VoiceAgent Skills article.
Building the VoiceAgent Knowledge Base
In this step, you provide the information your VoiceAgent can use during conversations. The Knowledge Base helps the agent answer caller questions based on the content you upload, such as PDFs, internal guides, FAQs, or other supporting documents.
A well-prepared knowledge base improves the quality and accuracy of the VoiceAgent's responses, especially for common questions and routine interactions.
Templates do not pre-populate the knowledge base. You will need to add your own sources after creation.
For detailed instructions on how to add, manage, and optimize knowledge sources, see the VoiceAgent Knowledge Base article.
Defining Scenarios and Guardrails
In the Scenarios step, you define how the VoiceAgent should behave in specific situations and what boundaries it should follow during conversations. Scenarios guide the agent's responses in common or important situations, while guardrails keep the agent within the intended scope.
This step helps make the VoiceAgent more reliable, predictable, and aligned with your business needs. For more detailed instructions, refer to the dedicated Scenarios and Guardrails article.
Testing the VoiceAgent
You can use the built-in test call option available during VoiceAgent creation or when editing an existing agent. This allows you to simulate a real call and verify how the VoiceAgent behaves before connecting it to your live call flow.
To start a test:
Click
Call to test in the testing panel during VoiceAgent creation, or click theTest VoiceAgentbutton in the top-right corner of the VoiceAgent edit page.A pop-up window will appear where you can enter the phone number the VoiceAgent should call. Enter your phone number in international format.
If your VoiceAgent requires any custom variables for the conversation, you will also be prompted to fill them in.
Click Start test call.
Once the test is started, the VoiceAgent will place a call to the number you provided. This allows you to verify how the agent behaves with the current configuration, including its greeting, conversation flow, and skills.
Pro tip: If your VoiceAgent is configured for inbound calls, you can also call the phone number where the VoiceAgent is included in the call flow to experience the interaction the same way your callers would.
Triggering VoiceAgent Calls
CloudTalk VoiceAgents can be used in two different ways, depending on when you want the AI to interact with your customers:
Inbound Calls
A VoiceAgent can handle incoming calls as part of your call flow. When a customer calls your number, the VoiceAgent can take over the conversation at a specific point in the flow. It can help with answering questions, checking order or ticket status, booking appointments, and more.
Creating a VoiceAgent does not automatically connect it to incoming calls. To make it answer inbound calls, you need to add it to the number’s Call Flow Designer.
To learn how to set this up, see the VoiceAgent for Inbound Calls guide.
Outbound Calls
You can also use VoiceAgent to make outgoing calls. These calls are triggered automatically using the CloudTalk API, often through tools like HubSpot, Zapier, Make, or your internal systems. This is useful for tasks like follow-ups, reminders, surveys, or proactive outreach without needing a human agent to start the call.
To learn more, see the VoiceAgent for Outbound Calls guide.
FAQ
What if I do not finish the VoiceAgent setup?
If the setup is not fully completed, the VoiceAgent may appear as incomplete and will not be fully ready for use until all required steps are finished.
Can I update the configuration later, or only during creation?
Yes. The four steps are used for the initial configuration, but you can edit everything afterward from the Agents list.
Can I customize a template after creating the agent?
Yes. Every part of a template-based agent can be edited after creation, including the goal prompt, greeting, skills, scenarios, and knowledge base. Templates give you a working starting point; you customize from there.
What should I prepare before starting?
It helps to know your use case in advance, including whether the VoiceAgent should handle inbound or outbound calls, what goal it should achieve, what information it should collect, and whether it needs knowledge documents or transfer rules. If you are not sure, start with a ready-made template and adjust from there.
Can I have different VoiceAgents for different phone numbers?
Yes. You can create multiple VoiceAgents, each with different skills, knowledge, and behavior, and assign them to different numbers using Call Flow Designer. One inbound VoiceAgent can also be used on multiple numbers if needed.
Can I see a transcript of what the AI actually said on a call?
Yes. Call recordings and transcripts are available in the call details page in case your account has an active Conversation Intelligence package.
What happens if the VoiceAgent transfers a call and nobody picks up?
The caller is not disconnected. VoiceAgent transfers to agents and groups are safe transfers: the VoiceAgent keeps the caller on the line while it reaches out to the target, and only hands the call over once the connection succeeds. If the agent doesn't answer or the group can't take the call, the caller stays with the VoiceAgent, which can offer alternatives such as taking a message. See the Transfer to Human skill guide for details.
Need help or have a question? Just reach out through our Support portal — we’re here for you.










