AI Receptionist in Webex Calling

AI Receptionist in Webex Calling enhances how operators and receptionists handle calls within your organization by introducing intelligence with the AI Receptionist. The AI Receptionist for Webex Calling acts as your virtual assistant, and is always available to help. It automates tasks such as answering calls, responding to simple questions, and transferring calls to a queue. With AI Receptionist automating day-to-day tasks, it helps improve your productivity and work quality.

The AI Receptionist is easy to deploy and simple to use. It integrates seamlessly with Webex Calling and supports transfers to individuals or Webex Calling resources.

AI Receptionist is accessed from Control Hub. You can customize how your virtual receptionist speaks, add key company information, and set up smart routing rules. Whether you’re supporting one location or many, AI Receptionist is flexible enough to match your business needs while ensuring every caller gets fast, intelligent service.

An Administrator can create, modify and manage the AI Receptionist agent for other users. You can add an AI Receptionist using these steps:

Assign a phone number— Choose a Webex Calling phone number where AI Receptionist will receive incoming calls. This is the number your customers dial to reach your business.
Set voice and language— Customize how AI Receptionist speaks to your callers. Select from natural-sounding voices to match your brand personality.
Add company information— Provide important business details like store hours, office locations, parking instructions, or holiday closures. AI Receptionist uses this information to answer routine questions, reducing your team’s workload.
Build a knowledge base— Upload documents containing FAQs, product info, or service policies. AI Receptionist scans and learns from this content to provide instant, accurate answers.
Create a welcome greeting— Use text to create a friendly, branded greeting that plays when customers call.

The AI Receptionist is available in select regions and currently supports English language for prompts and templates.

This feature isn’t available in India and Saudi Arabia.

Access AI Receptionist from Control Hub

The AI receptionist helps you manage front desk operations by answering calls and responding to common questions. It uses artificial intelligence to give callers a seamless experience and ensures every question is answered.

1	Sign in to Control Hub
2	Go to Calling under Services then click AI Receptionist
3	Click Create to create a new AI Receptionist to operate as a front desk executive and answer common questions.

Ordering and licensing

You can purchase AI receptionist agent as an add-on with Webex Calling professional license.

The trials have entitlement to 2 bundles of AI receptionist agent. Each entitlement provides 500 minutes of usage.
The usage is tracked in seconds, with the minimum consumption unit being 1 second.

The AI receptionist agent is provisioned in the Contact Center region that corresponds to the primary Webex Calling location for customers with only Calling, or to the Contact Center primary location for customers who have both Calling and Contact Center. Additionally, for understanding the Contact Center region mapping by country, you can refer to the Data Locality in Webex Contact Center article, which details the data center locations used for Webex Contact Center tenants based on the country of operation.

Create an AI Receptionist

To set up a new AI Receptionist, follow these steps:

1	Sign in to Control Hub
2	Go to Calling under Services then click AI Receptionist.
3	Select the AI Receptionist tab and click Create.
4	For the General Settings, enter the following: Location— Assign the receptionist to a specific location. For example: San Jose or Seattle. Name— Provide a unique name to identify the AI Receptionist. This name is used as the default Caller ID and reference the AI Receptionist. Phone Number/Extension— Assign a primary line and/or extension. Add a Primary Number and the Alternate numbers using the look up field. You can add up to 10 Alternate numbers, including numbers from multiple locations. AI engine— Webex AI Pro supports global languages and provides human-like interactions for most use cases. For scenarios requiring an enhanced human-like conversational experience, Webex AI Pro-US offers specialized, English-only support. Available AI engines are Webex AI Pro and Webex AI Pro-US. The Webex AI Pro-US engine is available only for organizations and locations in the United States region. Callers outside the United States cannot establish a call with an AI Receptionist that uses Webex AI Pro-US. Select the AI Language— Select the language from the drop-down menu. Example: English(United States) Select the AI Voice — Select the AI voice profile. Example: Jennifer. Direct line caller ID name— Set the caller ID name that is displayed when a user, virtual line, or workspace makes an external call. You can specify a custom name as the Direct line caller ID name, supporting Unicode characters up to 128 characters. Dial by name — Dial by name is used for auto attendant name dialing.
5	Configure the Receptionist Guidelines to allow the AI Receptionist to effectively communicate with the callers. Receptionist's Goal— Define the goals that the AI Receptionist must accomplish. For example: Help users book or cancel doctor appointments. Welcome Message— Enter the initial greeting. For example: Hello and welcome to Shine Healthcare clinic, how can I help you?. Receptionist guidelines— Provide specific behavioral rules, such as maintaining a professional tone or specific steps for booking tasks. Use the following templates to write instructions specific to your goal: `## 1. Identity` `-Role Definition:`—Define the persona and expertise of the AI Receptionist agent. For example, "You're Jamie, an expert customer service representative for any queries related to travel." `-Tone and Demeanor-`—Specify whether the agent should be friendly, formal, or casual. `##2. Context` `-Background Information`—Provide any necessary background details the agent should consider. For example, "This conversation is about booking travel for a family vacation." `-Environment Details`—Mention any system constraints such as the caller is calling over voice and may have background noise which may impact the quality of transcription. `##3. Task` `-Subtasks/Steps`—Break down the overall task into specific, sequential steps. For example, greeting, collecting travel dates, suggesting options, confirming details. Reference the actions at each step that will be used to fulfill the task. `-Optional step`—Additional information to handle specific tasks. For example, handle barge-in. `##4. Response Guidelines` `-Formatting Rules`—Define how to structure responses. For example, consider using bullet lists for options, clear numbering for steps in case of digital and short if there is voice. `-Language Style`—Provide instructions on formality, brevity, and clarity. `##5. Error Handling and Fallbacks` `-Clarification Prompts`—Define fallback questions when user input is ambiguous. For example, "I didn't catch that, could you please repeat your travel dates?" `-Default Responses`—Outline how the agent should respond if it can't process the request. For example "I'm sorry, I didn't understand. Can you try rephrasing?" `-Action Failures`—Provide guidelines for handling issues with actions integration with Webex Connect. `##6. User Defined Guardrails` `-Guardrail`—Remind the agent to keep the conversation restricted to the goal and not entertain any unrelated queries. `## 7. Examples` `-Sample Conversation`—Optionally add an example of the sample conversation between the end user and the AI Receptionist agent for better prompt adherence. Alternately, Click Apply template, to use a predefined template to configure your AI Receptionist. See Use the AI Receptionist templates for details.
6	Select or create a knowledge base to assign to the AI Receptionist. You can add files after creating the AI Receptionist. See Configure the Knowledge Base for details.
7	Set the Default action. The default action is triggered based on the caller's request or during specific error states. Play a message and end the call : You can choose to play either the default or custom message. Play default message Play custom message If the caller wishes to speak to an agent, select Transfer call. Transfer call: You can select a specific user or extension to act as the operator.
8	Click Review, to validate the settings for your new AI Receptionist. You can make changes, edits and manage the configuration by clicking the name of your AI Receptionist on the Features page.
9	Click Create to complete the configuration. A success message displays.

Edit an AI Receptionist

To modify how an AI Receptionist functions, follow these steps:

1	Sign in to Control Hub
2	Go to Calling under Services then click AI Receptionist.
3	Click on the AI Receptionist that you wish to modify.
4	From the General Settings, you can edit the following configurations: Slide the toggle to enable the AI Receptionist client. On enabling the toggle, AI Receptionist client is active. When the toggle is disabled, then the AI Receptionist client is inactive and will not answer any calls. Phone Number/Extension: Click the right arrow to expand the setting. You can assign a primary line and/or extension. Add a Primary Number and the Alternate numbers using the look up field. You can add up to 10 Alternate numbers, including numbers from multiple locations. Language and Voice: Select the AI Receptionist language (example: English US) and the AI Receptionist voice profile (example: John Doe). Caller ID: Enter the following: Direct line caller ID name: Set the caller ID name that is displayed when a user, virtual line, or workspace makes an external call. You can specify a custom name as the Direct line caller ID name, supporting Unicode characters up to 128 characters. Dial by name : Dial by name is used for auto attendant name dialing.
5	Configure the Receptionist Guidelines Agent's Goal: Define what the AI must accomplish. For example: Help users book or cancel doctor appointments. Welcome Message: Enter the initial greeting. For example: Hello and welcome to Shine Healthcare clinic, how can I help you?. Instructions for receptionist: Provide specific behavioral rules, such as maintaining a professional tone or specific steps for booking tasks. Use the following templates to write instructions specific to your goal: `## 1. Identity` `-Role Definition:`—Define the persona and expertise of the AI agent. For example, "You're Jamie, an expert customer service representative for any queries related to travel." `-Tone and Demeanor-`—Specify whether the agent should be friendly, formal, or casual. `##2. Context` `-Background Information`—Provide any necessary background details the agent should consider. For example, "This conversation is about booking travel for a family vacation." `-Environment Details`—Mention any system constraints such as the caller is calling over voice and may have background noise which may impact the quality of transcription. `##3. Task` `-Subtasks/Steps`—Break down the overall task into specific, sequential steps. For example, greeting, collecting travel dates, suggesting options, confirming details. Reference the actions at each step that will be used to fulfill the task. `-Optional step`—Additional information to handle specific tasks. For example, handle barge-in. `##4. Response Guidelines` `-Formatting Rules`—Define how to structure responses. For example, consider using bullet lists for options, clear numbering for steps in case of digital and short if there is voice. `-Language Style`—Provide instructions on formality, brevity, and clarity. `##5. Error Handling and Fallbacks` `-Clarification Prompts`—Define fallback questions when user input is ambiguous. For example, "I didn't catch that, could you please repeat your travel dates?" `-Default Responses`—Outline how the agent should respond if it can't process the request. For example "I'm sorry, I didn't understand. Can you try rephrasing?" `-Action Failures`—Provide guidelines for handling issues with actions integration with Webex Connect. `##6. User Defined Guardrails` `-Guardrail`—Remind the agent to keep the conversation restricted to the goal and not entertain any unrelated queries. `## 7. Examples` `-Sample Conversation`—Optionally add an example of the sample conversation between the end user and the AI agent for better prompt adherence.
6	Select or create a knowledge base for your AI receptionist. Manage files directly from the knowledge base page. See Configure the Knowledge Base for details.
7	Set the Default action. The default action is triggered based on the caller's request or during specific error states. Play a message and end the call : You can choose to play either the default or custom message. Play default message Play custom message If the caller wishes to speak to an agent, select Transfer call. Transfer call: You can select a specific user or extension to act as the operator.

Configure and manage intents

Intents help AI Receptionist understand when a caller wants to be transferred to a specific destination. The knowledge base helps AI Receptionist answer routine questions, while intents help it recognize the caller's purpose and route the call to the right person, group, or number.

For example, you can create intents for billing, sales, pediatric services, physical therapy, appointment scheduling, or any other service area that callers commonly request. Each intent includes a natural-language name, a description that explains when the intent applies, and a transfer destination.

AI Studio supports up to 9 intents and 1 default action for each AI Receptionist. The default action is used when AI Receptionist can't match the caller's request to an intent or when it reaches a configured fallback condition.

Before you begin, create an AI Receptionist and make sure the transfer destinations that you want to use are available in your organization.

1	Sign in to Control Hub
2	Go to Calling under Services, then click AI Receptionist.
3	Select the AI Receptionist that you want to configure.
4	Go to the Intents tab.
5	Click Add Intent.
6	In Intent name, enter a short name for the caller request that AI Receptionist must detect. The intent name must be unique for the selected AI Receptionist and can be up to 64 characters.
7	In Intent description, describe the caller request in natural language. Use clear language that explains when the call should be transferred. For example, enter Transfer callers who ask about physical therapy, rehabilitation, or injury recovery. The description can be up to 1024 characters.
8	Under Transfer to, choose a contact type. Users: Transfer the call to a user in your organization. Resources: Transfer the call to a Webex Calling resource, such as a call queue, hunt group, auto attendant, workspace, or virtual line. Organization contacts: Transfer the call to an organization contact. Number: Transfer the call to an external phone number.
9	Select the contact, resource, organization contact, or number for the intent.
10	Click Add Intent.
11	Repeat these steps until you add the intents that the AI Receptionist must handle.
12	To modify an intent, select the intent from the Intents list, update the name, description, or transfer destination, and click Save.
13	To delete an intent, select the intent from the Intents list, click Delete, and confirm the deletion.

When you delete an intent, AI Receptionist no longer transfers calls for requests that matched that intent. If a caller asks for that service again, AI Receptionist uses another matching intent or the configured default action.

Use the AI Receptionist templates

AI Receptionist templates are inbuilt examples designed to help you build receptionist clients on the platform. These templates come with pre-configured goals, messages, and instructions that demonstrate core features and best practices, serving as hands-on guides for understanding the feature's capabilities. Use them to explore the AI Receptionist creation process, experiment with configurations, and develop the skills needed to build your own custom clients effectively. These are industry-specific templates that can serve as a starting point, allowing you to customize the client to meet your specific requirements.

To use a template, lets consider the use case of booking an appointment at the doctor's office.

1	Sign in to Control Hub
2	Go to Calling under Services then click AI Receptionist.
3	Select the AI Receptionist tab and click Create.
4	For the General Settings, enter the following: Location— Assign the receptionist to a specific location. For example:France. Name— Provide a unique name to identify the AI Receptionist agent. This name is used as the default Caller ID and reference the AI Receptionist. Phone Number/Extension— Assign a primary line and/or extension. Add a Primary Number and the Alternate numbers using the look up field. You can add up to 10 Alternate numbers, including numbers from multiple locations. AI engine— Webex AI Pro supports global languages and provides human-like interactions for most use cases. For scenarios requiring an enhanced human-like conversational experience, Webex AI Pro-US offers specialized, English-only support. Available AI engines are Webex AI Pro and Webex AI Pro-US. The Webex AI Pro-US engine is available only for organizations and locations in the United States region. Internal and external callers must be physically located in the United States to connect to an AI Receptionist that uses this engine. Callers outside the United States cannot establish a call with an AI Receptionist that uses Webex AI Pro-US. Select the AI Language— Select the language from the drop-down menu. Example: English(United States) Select the AI Voice — Select the AI voice profile. Example: Jennifer. Direct line caller ID name— Set the caller ID name that is displayed when a user, virtual line, or workspace makes an external call. You can specify a custom name as the Direct line caller ID name, supporting Unicode characters up to 128 characters. Dial by name — Dial by name is used for auto attendant name dialing.
5	Configure the Receptionist Guidelines to allow the AI Receptionist to effectively communicate with the callers. Click Apply Template to select from the predefined rules. The drop down displays the available inbuilt template that can be selected based on the intent of the AI agent. This template demonstrates an autonomous AI Receptionist for managing doctor appointment bookings. This client contains actions to check availability of slots, create appointments, lookup appointments, and check insurance details.
6	Select or create a knowledge base to assign to the AI Receptionist. You can add files after creating the AI Receptionist. See Configure the Knowledge Base for details.
7	Set the Default action. The default action is triggered based on the caller's request or during specific error states. Play a message and end the call: Select a message to play from the following: Play default message Play custom message If the caller wishes to speak to an agent, select Transfer call. Transfer call: You can select a specific user or extension to act as the operator.
8	Click Review, to validate the settings for your new AI Receptionist. You can make changes, edits and manage the configuration by clicking the name of your AI Receptionist on the Features page.
9	Click Create to complete the configuration. A success message displays. A success message displays. You can add files to the knowledge base to facilitate the AI receptionist to effectively communicate with callers.

Supported locales and voices for the AI agent

The following table lists the English locales and voices supported for AI Receptionist when you use the Webex AI Pro or Webex AI Pro-US engine.

Table 1. Supported English locales and voices
Language	Locale	AI engine	Voice name
English	en-US	Webex AI Pro 1.0	en-US-Maria en-US-Daniel en-US-Jennifer en-US-Henry en-US-Sophia en-US-Ezra en-US-Hazel en-US-Oliver en-US-Clara
English	en-US	Webex AI Pro-US 1.0	en-US-Jess en-US-Lisa en-US-Frank en-US-Mia en-US-Chris en-US-Aaron
English	en-IN	Webex AI Pro 1.0	en-IN-Ananya en-IN-Veer en-IN-Arjun en-IN-Aarav en-IN-Kavita
English	en-GB	Webex AI Pro 1.0	en-GB-Agnes en-GB-Colton en-GB-Emma en-GB-John en-GB-Charlotte
English	en-GB	Webex AI Pro-US 1.0	en-GB-Lillian
English	en-AU	Webex AI Pro 1.0	en-AU-Amelia en-AU-Chris en-AU-Mia en-AU-Henry en-AU-Olivia
English	en-IE	Webex AI Pro 1.0	en-IE-Liam en-IE-Fiona
English	en-PH	Webex AI Pro 1.0	en-PH-Angela en-PH-James
English	en-SG	Webex AI Pro 1.0	en-SG-Emma en-SG-Lucas

Configure the Knowledge Base

The Knowledge Base (KB) provides the intelligence for the AI Receptionist to answer questions. It contains the documents and information that the AI Receptionist agent uses to answer the caller’s questions.

1	To navigate to the Knowledge Base, go to Calling > AI Receptionist > Knowledge Base
2	To create a new Knowledge Base, click Create.
3	Enter the following: Name: Enter a name for the Knowledge Base Description: Provide a description on the usage of the Knowledge Base. For example: Billing and insurance guidelines Add: You can either upload or create the document that the AI Receptionist can refer. Alternatively, you can drag and drop the files to the knowledge base. Alternatively, click Add File to add a file. The system supports the following file formats: PDF, .docx, .doc, .txt, .xlsx, .xls, and .CSV. The following file upload restrictions apply: Total storage limit for all files combined across multiple KBs: 2 GB Maximum number of files: 100 files per KB. Individual file size limit: 10 MB. Individual text (.txt) file size limit: 2 MB. PDF file size limit: In addition to the file size limit, a limit of 300 pages per PDF is also applicable. To ensure optimal processing when uploading Microsoft Excel (.xlsx) files, please follow these guidelines: The headers must be in the first row for the system to process the header row accurately. The system merges all tables within a single sheet into one. Therefore, clean your file to remove unnecessary data before uploading. Empty cells are automatically excluded from the ingestion process. Excel processing is more resource-intensive than other formats. To improve performance, split large spreadsheets into smaller files, keeping each file under 40 KB. Hyperlinks are supported in .docx, .xlsx, and .xls files. Note that PDF and CSV formats do not support hyperlinks. The system reads data from all sheets, including hidden ones. If you do not want specific content ingested, please delete it from the file before uploading. Create: Manually type or paste relevant text directly into the Control Hub. Download: The AI Receptionist Knowledge Base lets you track and download the uploaded files. Click the download icon to retrieve the documents that the Knowledge Base uses to answer queries, so you can review or update them when needed.
4	On creating the Knowledge Base, Assign or link it to an AI Receptionist during or after its creation.
5	Specify the Default Action. You must configure what happens if the AI Receptionist cannot answer a query. The following options are available: Play a message and end the call: Upload or record a custom .WAV file. Transfer call: Select a specific user or extension to act as the operator.
6	Review or verify the configuration and click Add.

Manage AI Receptionist

You can monitor and modify your AI Receptionist agents from the main dashboard.

Edit Settings: Click an existing AI Receptionist from the list to update the phone number, language, guidelines, knowledge base, or default action.
Enable/Disable: Use the toggle within the receptionist's settings to turn the service on or off. When the toggle is disabled, the AI Receptionist is inactive and doesn't answer calls.
Intents: Add, edit, or delete intents that help AI Receptionist route callers to the right user, Webex Calling resource, organization contact, or number. See Configure and manage intents.
Monitoring Usage: Open the Usage tab to view entitlement, consumed usage, remaining usage, and billing-cycle details for AI Receptionist. See View AI Receptionist usage.
Reports: Use the Calling AI Receptionist Stats report to understand call volume, answered calls, transfer attempts, transfer success, intent transfers, default transfers, and talk time. See AI Receptionist reports.
Search and Filter: Use the search bar or location filter to quickly find specific receptionists in large organizations.
Modify phone numbers: You can modify phone numbers within AI Receptionist and maintain alternate numbers.
Delete: If a receptionist client is no longer needed, select Delete from the list.

When you delete an AI Receptionist agent, the associated Knowledge Base is still available for reassignment and use in other AI Receptionists.

View AI Receptionist usage

The Usage tab helps administrators track how much AI Receptionist entitlement is available and how much has been consumed during the current billing cycle. Use this information to understand adoption, monitor consumption, and plan whether you need more entitlement before callers are affected.

Usage details are shown for the AI Receptionist feature in Control Hub and are refreshed from the usage and billing backend for the billing cycle.

1	Sign in to Control Hub
2	Go to Calling under Services, then click AI Receptionist.
3	Click the Usage tab.
4	Review the usage summary for the current billing cycle. The usage view can include the total entitlement, consumed usage, remaining usage, and the billing cycle for which usage is displayed.
5	Use the usage summary to identify whether the organization is approaching its entitlement limit.
6	If usage is higher than expected, review the configured AI Receptionists, knowledge bases, intents, and default actions to confirm that callers are being handled and transferred as intended.

Usage values are intended for monitoring and billing-cycle awareness. For call handling details, use the AI Receptionist report.

AI Receptionist reports

The AI Receptionist report helps administrators understand how AI Receptionist is handling calls across the organization. Use the report to see whether callers are reaching AI Receptionist, whether calls are being answered, how often calls are transferred, and whether transfers succeed.

The report is useful when you want to evaluate front desk automation, compare performance across receptionists or locations, and identify configuration improvements. For example, a low intent-transfer rate may indicate that intents need clearer descriptions, while a low transfer success rate may indicate that transfer destinations, operating hours, or routing configuration need review.

Administrators can generate the Calling AI Receptionist Stats report on demand or schedule it from Control Hub.

What you can infer from the report

A higher percentage of calls directly handled by AI Receptionist compared to the total calls presented indicates that AI Receptionist is effectively managing a larger portion of the incoming calls, demonstrating its efficiency in handling customer interactions.
High default transfers can indicate that caller requests are not covered by the knowledge base or by configured intents.
Low intent-transfer usage can indicate that intent names or descriptions need to be clearer or that callers are asking for services that are not configured as intents.
Low transfer success rate can indicate a problem with transfer destinations, operating hours, external number permissions, or downstream routing.
Increasing total talk time or average talk time can indicate higher caller demand, unclear knowledge base content, or callers needing more assistance before transfer.

Report KPIs

KPI	Description	Format	Minimum/maximum value	Example
AI Receptionist	Name of the AI Receptionist. Use this to identify which receptionist handled the calls.	Text	Not applicable; based on configured name.	Shine Healthcare Clinic
Phone Number	Primary phone number assigned to the AI Receptionist.	Phone number, usually E.164 format	Not applicable; based on assigned number.	+14085550100
Extension	Extension assigned to the AI Receptionist, if configured.	Number string	Not applicable; based on assigned extension.	3991
Location	Location associated with the AI Receptionist. Use this to compare activity by site.	Text	Not applicable; based on configured location.	San Jose
Total Presented Calls	Total number of presented calls that AI Receptionist answered. Compare this with Calls Presented to identify calls that were not answered.	Whole number	Minimum 0; maximum can't exceed Calls Presented for the selected report scope.	1195
Total directly handled calls	The total number of calls presented that were directly handled by the AI Receptionist. This means that the calls terminated at AI Receptionist without any transfers and is a key indicator of efficiency.	Whole number	Minimum 0; no fixed maximum.	1250
Total Transfers Attempted	Total answered calls where AI Receptionist attempted a transfer. This includes intent transfers and default transfers.	Whole number	Minimum 0; maximum can't exceed Calls Answered for the selected report scope.	640
Total Successful Transfers	Total calls that were successfully transferred. This includes successful intent transfers and successful default transfers.	Whole number	Minimum 0; maximum can't exceed Transfers Attempted.	602
Percentage Transfer Attempt	Percentage of answered calls where AI Receptionist attempted a transfer. Use this to understand how often callers need a live destination.	Percentage	0% to 100%	53.6%
Percentage Transfer Success	Percentage of attempted transfers that completed successfully. Use this to check whether transfer destinations and routing are working as expected.	Percentage	0% to 100%	94.1%
Attempted Intent Transfers	Total calls where AI Receptionist attempted a transfer based on a matched intent. Use this to measure intent usage.	Whole number	Minimum 0; maximum can't exceed Transfers Attempted.	420
Successful Intent Transfers	Total calls successfully transferred based on a matched intent. Use this to understand whether intent routing is successful.	Whole number	Minimum 0; maximum can't exceed Intent Transfers Attempted.	398
Attempted Default Transfers	Total calls where AI Receptionist attempted a transfer by using the default action. Use this to understand fallback transfer usage.	Whole number	Minimum 0; maximum can't exceed Transfers Attempted.	220
Successful Default Transfers	Total calls successfully transferred by using the default action. Use this to validate fallback routing.	Whole number	Minimum 0; maximum can't exceed Default Transfers Attempted.	204
Total Talk Time	Total time AI Receptionist spent talking to callers. Use this to understand total handled conversation time.	The time format is in HH:MM:SS.	Minimum 0; no fixed maximum.	14:35:30
Average Talk Time	Average time AI Receptionist spent talking to callers per answered call. Use this to understand whether calls are brief, lengthy, or changing over time.	The time format is in HH:MM:SS.	Minimum 0; no fixed maximum.	00:01:48

Use the report together with usage information. Usage shows entitlement consumption for the billing cycle, while the report shows operational call handling and transfer effectiveness.

Known limitations and considerations

Read these pointers to know the limitations while using the AI Receptionist:

Impact of Announcements on AI Receptionist

Mandatory system announcements such as compliance and call recording disclosures may interrupt the AI Receptionist (AIR) welcome prompt for organizations in the EUN region. This can result in the caller hearing an incomplete welcome message.

Use case: Create AI Receptionist intents for a hospital

This use case provides an example of using AI Receptionist as a hospital front desk assistant. The AI Receptionist answers common caller questions and uses intents to transfer callers to the right hospital department, such as pediatrics, cardiology, billing, pharmacy, physical therapy, or general appointments.

In this example, each intent represents a department or service area. The intent description tells AI Receptionist when the caller should be transferred, and the transfer destination maps the intent to a user, Webex Calling resource, organization contact, or number.

AI Studio supports up to 9 intents and 1 default action for each AI Receptionist. Create intents for the departments that receive the highest caller volume, and use the default action for unmatched requests or fallback scenarios.

Go to Calling under Services, then click AI Receptionist.

Select the AI Receptionist tab and click Create.

For the General Settings, enter the hospital front desk AI Receptionist information.

Location: Assign the AI Receptionist to the hospital location. For example, San Jose Hospital.
Name: Enter a name that identifies the AI Receptionist. For example, Main Hospital Receptionist.
Phone Number/Extension: Assign the main hospital phone number and extension where callers reach the AI Receptionist.
Language and Voice: Select the language and AI voice profile for the receptionist.
Direct line caller ID name: Set the caller ID name that represents the hospital reception line.
Dial by name: Configure dial by name if callers must be able to reach staff by name.

Configure the Receptionist Guidelines so that AI Receptionist communicates clearly with hospital callers.

Receptionist's Goal: Define the goal. For example, answer routine hospital questions and transfer callers to the correct department based on their request.
Welcome Message: Enter the greeting. For example, Welcome to City Hospital. How can I direct your call today?
Receptionist guidelines: Add instructions to keep the conversation professional, ask clarifying questions when the requested department is unclear, and avoid providing medical advice.

Select or create a knowledge base to assign to the AI Receptionist.

The knowledge base can include hospital hours, parking details, visitor policies, department descriptions, appointment information, billing information, and pharmacy hours.

Set the Default action.

The default action is used when AI Receptionist can't match the caller request to one of the configured department intents or when it reaches a fallback condition.

For a hospital, you can configure the default action to transfer callers to the main operator or play a message and end the call.

Click Review to validate the AI Receptionist settings.

Click Create to complete the AI Receptionist configuration.

After the AI Receptionist is created, go to the Intents tab and create intents for the hospital departments.

The following table shows example department intents that you can create.

Intent name	Intent description	Transfer destination
Pediatric Services	Transfer callers who ask about child health, pediatric appointments, vaccinations, or pediatric care.	Pediatrics call queue
Cardiology	Transfer callers who ask about heart care, cardiology appointments, cardiac tests, or cardiology results.	Cardiology hunt group
Billing and Insurance	Transfer callers who ask about medical bills, insurance coverage, claims, payment, or account balances.	Billing department user or call queue
Pharmacy	Transfer callers who ask about prescriptions, refills, medication pickup, or pharmacy hours.	Pharmacy phone number
Physical Therapy	Transfer callers who ask about rehabilitation, injury recovery, therapy appointments, or physical therapy services.	Physical Therapy call queue
General Appointments	Transfer callers who want to schedule, reschedule, or cancel a hospital appointment.	Appointments auto attendant

For each department intent, enter the intent name, intent description, and transfer destination, then click Add Intent.

Use clear natural-language descriptions so AI Receptionist can match caller requests accurately. For detailed intent configuration, see Configure and manage intents.

When callers reach the hospital AI Receptionist, it can answer common questions from the knowledge base and transfer callers to the appropriate department when their request matches a configured intent. If no intent matches, the configured default action is used.

Troubleshooting known issues

Here are few common error scenarios and troubleshooting tips:

Error	Resolution
Location Errors	If you see a message stating Cannot create AI Receptionists for this location, ensure the location is properly configured for Webex Calling.
Fallback Responses	Always configure a Default Action to ensure callers are not left without assistance if the AI encounters an unknown request.
Guardrails	Use the Instructions section to set strict boundaries, such as Do not provide medical advice, to ensure compliance and safety.
Connection Failures	If the connection to the Cloud Contact Center fails, the system automatically falls back to the configured Default Action.
Monitoring	Administrators can track the session in minutes using the Analytics in Control Hub.

Cisco AI Assistant

Thanks for your feedback.