AI Agent Studio

AI Agents are designed to automate conversations and improve customer interactions. By leveraging advanced artificial intelligence, AI Agents seamlessly resolve issues and perform actions with human-like decision-making, resulting in faster resolutions and happier customers. Benefits of AI Agents for Customers:24/7 customer support: Ensure queries are handled at all times, enhancing customer satisfaction and loyalty, especially in retail and e-commerce environments.Reduced wait times:  Streamline processes like ordering to minimize delays.Multilingual support: Communicate with customers in multiple languages, allowing you to support a global customer base.Multi-channel availability to provide support across chat, email, SMS, WhatsApp, Kustomer Voice, and forms....and more!Who can access this feature?User typesAdmins can access the AI Agent Studio page.In this article:What powers AI Agents?AI Agent TeamsAI AgentsSupervisor agentToolsKnowledgeObservabilityGuardrailsReportingBest practice for disclosure of an AI botThis article introduces the main concepts you need to understand and build an AI Agent team for your organization. To learn more about configuring an AI Agent team, see our Kustomer University course.What powers AI Agents?AI Agents are powered by our reasoning engine, which thinks like a human to make decisions. It adapts to new information customers provide, handling uncertainty and generating dynamic responses based on changing contexts.AI Agent teamsAI Agent teams are customizable to suit your organization's needs. When creating a team, consider what problems or challenges you want the agents on that team to address. You may also want to create teams based on the channels they’ll respond to, such as chat, email, or SMS, to ensure tailored responses for each channel. If your organization manages multiple brands, you can also create a separate, dedicated team for each brand, ensuring that each one maintains its unique voice and tone while responding to inquiries related to its products or services. How AI Agents workAI Agents are autonomous and intelligent systems that use Large Language Models (LLMs) to help answer customer questions and solve problems. They work together as a team, alongside your human agents, to resolve customer inquiries.AI Agents work in an iterative cycle, meaning they process new information as it comes in and adjust their actions as needed. They use tools to interact with the system and take actions based on the conversation. These actions can include things like looking up customer details or updating order information both inside and outside of Kustomer.When setting up an AI Agent, you decide its job, how it should respond, and what tools it can use. For example, let's say you are setting up an AI Agent for an airline's customer support call center. These agents would interact with reservation databases and ticketing systems found in your call center system. The AI Agents can then use tools to perform actions such as changing flight bookings or handling cancellations.Supervisor agentWhen you create a new AI agent team, a "supervisor" agent is automatically added. This agent greets customers and is the only one that talks to them directly. It can access tools and your knowledge base to perform needed actions and provide helpful answers. The supervisor agent also works with other AI agents on the team to gather the necessary information and take appropriate actions to resolve customer issues.ToolsEach AI Agent should be equipped with the tools to take relevant actions and resolve customer inquiries. Examples of tools an AI agent can use include access to customer data, product inventory, or order history. You control which tools an agent can use and when ensuring they have the right resources to assist customers. This tutorial provides examples of AI Agents and tools you can consider using to respond to e-commerce inquiries.KnowledgeAI Agents can leverage your organization’s public knowledge base(s) to respond accurately to customer inquiries. By tapping into your existing knowledge base, AI Agents enhance the consistency of your customer support. They can quickly locate the most relevant information, reducing the time it takes to resolve issues and providing customers with clear and accurate responses. Learn how to build a team that uses your knowledge base to respond to customer inquiries in this tutorial.ObservabilityThe Traces page provides detailed logs of interactions between AI Agents and customers. Traces help you monitor an AI Agent team's performance, understand its decision-making, and identify areas for improvement.Traces include details such as:Timestamps of each step in the interaction.Inputs received from the customer, such as messages.The name of the tool that was used to respond.Outputs sent back to the customer, such as replies.The help article used to respond, if applicable. Learn more about observability and how to read a trace in this article.GuardrailsGuardrails are crucial for ensuring that your AI agents adhere to company policies, particularly when it comes to sensitive topics like competitors. By setting guardrails, you can prevent your AI agents from mentioning or discussing competitors in customer interactions. In addition to protecting against mentions of competitors, these guardrails work alongside standard safeguards within the product to ensure AI agents don’t disclose sensitive information. Guardrails can also be used strategically to amplify your products and services' benefits and unique differentiators. By setting these guidelines, you help your AI agents maintain focus on brand integrity and consistently communicate the value your products and services bring to customers.ReportingThe AI Agents standard report lets you track the quality of responses generated by AI Agents and track key metrics such as the number of conversations they handle. This report helps you evaluate the effectiveness of your AI Agents and identify opportunities to optimize their performance.Best practice for disclosure of an AI botWhen required by applicable law and to provide transparency to customers, it should be made clear that they aren't interacting with a human and that the person is interacting with an AI bot. While it should still feel like they are having a friendly conversation with a machine, but it must be made clear to customers that they are communicating with an AI bot:at the beginning of any conversation or message thread,after a significant lapse of time, orwhen a conversation moves from human interaction to automated experience.Acceptable disclosures include but are not limited to: “Hello! Welcome to [Company name]. I am an AI bot and am not a real person, but I’ll do my best to answer your questions.”“Hello! I’m the [Company Name] AI bot,”“Hello! You are interacting with an AI bot,” “Hello! You are talking to a AI bot,” Note: Even where not legally required, we recommend informing users when they’re interacting with an AI bot as best practice, as this helps manage user expectations about their interaction with your messaging experience. We recommend that you consult with your legal counsel to ensure that you are complying with all applicable laws regarding the use of AI Agents.You can customize this message for an AI Agent team responding through the chat or email channel by going to the Settings tab for that team.

E-commerce organizations can use AI Agent Teams to significantly streamline customer inquiries, especially when dealing with high volumes of questions related to orders and returns. This tutorial will provide guidelines on the types of AI Agents and tools you may want to set up to handle these inquiries, focusing on specific areas such as order tracking and return management. By organizing agents around these critical functions, your business can improve response times and deliver more personalized, efficient support.Who can access this feature?User typesAdmins can access the AI Agent Studio page.In this article:Example AI Agents for your teamExample toolsExample AI Agents for your teamWhen setting up an AI Agent Team with multiple AI agents, it’s important to define clear roles for each one to ensure they can efficiently handle specific customer inquiries. In an e-commerce setting, you may want to create agents specializing in different customer service areas, such as order inquiries or managing returns. Below are examples of the types of agents you might want to consider for your e-commerce team.Note: The following agents and prompts are only examples. The exact prompts and agents used will vary depending on your business and goals.SupervisorThe Supervisor is the main point of contact for customers, managing inquiries and resolving issues. The supervisor is the only agent who communicates directly with customers. It can collaborate with other AI Agents on the team to gather the necessary information to address customer questions.Example prompt:You work for a company called [X]. You are a customer service representative working on a team, and you are the customer-facing agent. Your JOB RESPONSIBILITIES and TEAMMATES are below: ## JOB RESPONSIBILITIES - Kindly greet the customer - Determine the customer's intent - Address the customer's issue immediately. If you need assistance, ask your TEAMMATES for assistance immediately. - Mark the Conversation as Done with the MarkConversationDone tool when the customer has confirmed they no longer need any help - Do not ask the customer for information that you can get on your own, or via one of your tools or teammates - Do not attempt to multi-task or put the customer on hold ## TEAMMATES - OrderReturnExpert: A teammate that is an expert on order returns. You MUST use them for anything involving order returns - OrderRefundExpert: A teammate that is an expert on order refunds. You MUST use them for anything involving order refunds - OrderSalesExpert: A teammate that is an expert on sales and placing orders. You MUST use them for anything involving order salesOrderSales agentThis AI Agent handles all sales inquiries. For example, this agent could help answer an inquiry when a customer wants more information about your products or wants to place an order.Example prompt:You work for a company called [X] in the customer service department. You have a teammate that will consult you when a customer would like information on our products, or placing an order. Your JOB RESPONSIBILITIES and OPERATING PROCEDURES are below: ## JOB RESPONSIBILITIES * Provide answers to your teammates to the best of your ability * Help customers place orders using the Order Placement Operating Procedure ## OPERATING PROCEDURES ### Sensitive Payment Information 1. Look up if the customer has existing Payment Information 2. Mask the Payment Information in all communications. ### Order Placement. You MUST do all of the following when placing orders for customers: 1. Confirm which product the customer would like to order. If they are not sure, help them identify which product would fit their needs. 2. Determine which payment method the customer would like to use for the order. You MUST use the Sensitive Payment Information Operating Procedure before continuing. 3. Confirm the product and price with the customer once more, stating the price and the product you will be ordering on their behalfOrder Return Expert agentThis AI Agent is an expert at handling order returns and should be contacted when a customer wants to initiate a return for an item.Example prompt:You work for a company called [x] in the customer service department. You have a teammate that will consult you when a customer needs help with an order return. Your JOB RESPONSIBILITIES and OPERATING PROCEDURES are below: ## JOB RESPONSIBILITIES * Provide answers to your teammates to the best of your ability * Handle all return inquiries using the tools and information provided to you ## OPERATING PROCEDURES Below are specific instructions to handle certain common scenarios ### Returns 1. Make sure a customer does not already have a Return for the relevant order. If they do, do not proceed 2. Create a Return for the Order that is being returned using the CreateReturn tool. You may need to fetch the Customer's orders first to complete this 3. Mark the Order as Returned using the MarkOrderAsReturned toolOrder Refund Expert agentThis AI Agent is an expert in order refunds and should be contacted when a customer wants to initiate a refund or a change in their order requires them to receive a refund.Example prompt:You work for a company called [X] in the customer service department. You have a teammate that will consult you when a customer needs help with an Order refund. Your JOB RESPONSIBILITIES and OPERATING PROCEDURES are below: ## JOB RESPONSIBILITIES * Provide answers to your teammates to the best of your ability * Handle all order refund inquiries using the tools and information provided to you ## OPERATING PROCEDURES Below are specific instructions to handle certain common scenarios ### Refunds When a customer wants a refund: 1. Check that the customer does not have a Refund Request for the Order already using the GetRefundRequests tool. If they do, do nothing. 2. Create a Refund Request with the relevant information using the CreateRefundRequest tool. You may need to use the GetCustomerOrders tool to find the orderId. 3. Create a Return for the Order using the CreateReturn tool 4. Ask the Refund Reviewer to review the Refund Request. This is a critical step that MUST be done. No refund can be processed unless the Refund Reviewer reviews it, so you MUST do this."Customer Data Expert agentThe Customer Data Expert is an expert on all customer data and can retrieve customer data such as orders, refunds, returns, emails, etc.Example prompt:You work for a company called [X] in the customer service department. You have a teammate that will consult you when they need customer data. Your JOB RESPONSIBILITIES are below: ## JOB RESPONSIBILITIES * Provide answers to your teammates to the best of your ability, using the tools and information provided to you"Refund reviewer agentThe Refund Reviewer Agent is responsible for reviewing and approving refund requests. Once they approve a refund request, it is sent for processing. The Refund Reviewer might be used when refund requests are created.Example prompt:"You work for a company called [X] in the customer service department as a Refund Reviewer. You have a teammate that will consult you when a Refund Request has been created and is ready for review. Your JOB RESPONSIBILITIES and OPERATING PROCEDURES are below: ## JOB RESPONSIBILITIES * Review Refund Requests, and either Approve or Deny them based on the Refund Approval Process ## OPERATING PROCEDURES Below are specific instructions to handle certain common scenarios ### Refund Approval Process 1. Make sure a customer does not already have a Refund for the relevant order. If they do, do not proceed 2. Ensure that the Refund would conform with the company's refund process 3. Mark the Refund Request as Approved or Denied, with an appropriate Review DetailExample toolsEquipping your AI agents with the right tools is crucial to ensuring they can respond effectively to customer inquiries. Depending on the agent’s role, each will need access to specific data, such as order records, shipping information, or customer details. By assigning the appropriate tools, your AI agents can quickly provide accurate and helpful responses. Below are examples of the tools your agents may need and which agent(s) should have access to them.ToolDescriptionAgent(s) that can accessMark Conversation DoneMarks the conversation as done once the inquiry has been resolved.Support Team SupervisorCreate OrderCreates a new order for a customer.Order Sales agentGet Customer InfoGets a customer's information details.Order Sales agentOrder Return ExpertCustomer Data ExportCreate ReturnCreates a new return request for a customer.Order Return ExpertCreate Refund RequestCreates a new refund request for a customer.Order Return ExpertMark Order as ReturnedMarks an order as returned once a refund request is created.Order Return ExpertGet Customer OrdersGets a customer's order details.Order Return ExpertCustomer Data ExpertGet Refund RequestGets details from a refund requestOrder Return ExpertCustomer Data ExpertRefund Reviewer AgentGet Customer Contact InfoGets a customer's contact information such as, email or phone number.Customer Data ExpertApprove Refund RequestApproves a refund request. Refund Reviewer AgentDeny Refund RequestDenies a refund request.Refund Reviewer Agent

In this tutorial, you'll learn how to create a team with a single AI Agent that uses your knowledge base or external URLs to respond to conversations across all or selected communication channels. Who can access this feature?User typesAdmins can access the AI Agent Studio page.Learn more about AI Agents in this article.In this articlePrerequisitesCreate your AI Agent teamDefine your AI Agent's roleSpecify conditions and deploy your teamPrerequisitesBefore creating this AI Agent team:You must have a public knowledge base.You must create a data source.Create your AI Agent teamThe first step is creating a team to respond to these inquiries. Go to Settings  > KIQ > AI Agent Studio and select Add AI Agent Team.Enter a name and brief description of what this team will do. In this example, we are creating a team that will answer general company FAQ's.Select Create.Define your AI Agent's roleNext, we will define your agent's role on this team. By default, every newly created team has a single agent already created whose purpose is to be the first AI Agent your customers will interact with. On the team page you just created, go to the AI Agent Team tab and select the existing Supervisor role.Enter a short description of this agent's role on your team.Select the Knowledge base and public URL Data Source the supervisor agent should use to respond to customers from the drop-down menu.Then, enter the instructions the supervisor agent will follow to answer an inquiry. When writing instructions, be specific about how you want the AI to respond. Here is an example prompt we can use.You work for a company called [X]. You are a customer service agent for [X]. Your JOB RESPONSIBILITIES and OPERATING PROCEDURES are below: ## JOB RESPONSIBILITIES - You respond to general inquiries about our company for topics such as our shipping and handling policy and return timeline. Other examples of general questions you can answer are about topics such as our business hours and how customers can contact us. ## OPERATING PROCEDURES - You only use information from our public knowledge base articles and the public URLs provided to answer customer questions. - Once you provide the answer, you will ask the customer if there is anything else you can help with. - If the customer replies with an answer similar to "no" or doesn't have any other questions, respond with "Thanks, have a great day!" - If the customer responds with an answer similar to "yes" or responds with additional questions, continue to answer their questions using the information from our public knowledge base articles or public URLs.Select Save Changes.Specify conditions and deploy your teamLastly, go back to your team page and select Deploy to set specific conditions for this team to respond to inquiries. You can set conditions like customer email, company, or even custom attributes like their tier status to ensure the AI agent responds only when criteria are met. This flexibility helps tailor responses to meet your organization’s needs.Note: If you do not set a specific channel, the AI Agent team will respond to inquiries from any supported channels.Once you set any applicable conditions, select Deploy again to make your changes to live. Doing so automatically activates the AI Agent team. The team will respond to all incoming messages that match the set conditions.You can select Observe or View Traces  from the main AI Agent Teams page at any time to see a detailed log of the team's interactions with customers.

Observability refers to the ability to monitor, track, and understand the performance of your AI Agent teams. The key feature of our observability tool is traces, which provide a detailed view of how the AI Agent processes an inbound message from start to finish. Traces allow you to see the journey a request takes through different parts of the AI system, giving you a granular look at each step in the process.By analyzing traces, you can gain insight into how effectively an AI Agent Team handles customer interactions, identify issues, and make improvements to ensure optimal performance.Who can access this feature?User typesAdmins can access the AI Agent Studio page.Learn more about AI Agents in this article.In this articleWhat are traces?View traces for an AI Agent teamHow to read a traceUnderstand guardrailsWhat are traces?Traces follow a request's lifecycle as it moves through your AI Agent. From the moment an inbound message is received, traces log each interaction, decision point, and operation the AI performs. This includes accessing tools and retrieving or updating data in Kustomer.Each trace provides a timestamped breakdown of the following:Any articles in your public knowledge base that were used to respond to the inquiry.Reasoning operations provide insight into how the AI arrived at a decision or why it took a particular action while handling a customer inquiry.The message each AI Agent received from the supervisor.Checks to ensure that both default and custom guardrails are followed.View traces for an AI Agent teamYou can view traces for both AI Agent teams that are currently deployed and past versions of a team. From the AI Agents Studio page:Go to Settings > Kustomer IQ > AI Agent Studio.Select View Traces for the team whose performance you want to analyze. By default, you will see a list of traces for all conversations answered by the latest deployed team.From directly within a conversation:Go to a conversation answered by an AI Agent and select Conversation options.Then, select Observe traces.You can select any trace to see details of what took place for every inbound message in the conversation.How to read a traceNote: The results you see in your traces may vary based on how you have your AI Agent team set up.A trace shows all the operations that occurred to generate the final response the AI sent to the customer. You can select each operation to see details, such as which help article was used and how the AI reasoned its response.In this section, we will walk through the sample trace below to see what information you can gather from this feature.1. E-Comm Support Supervisor responseYou can start to review a trace by selecting the Supervisor agent at the top. Here, you can see that the inbound message from the customer was "I would like to return the popcorn bowl." and that the AI responded with:Hi John! I've initiated the return process for the popcorn bowl. Could you please let me know the reason for your return? This will help me assist you better!2. Knowledge searchOnce an inbound message comes in, the AI performs a Knowledge Search operation to see if there is a help article that can answer the inquiry. If help articles are found, you will see the article's name, the content the AI extracted from it, and a direct link to it.3. ReasoningThe Reasoning operation gives you visibility into the AI’s internal decision-making process and helps you understand why the AI did what it did. Here, you can see that the AI decided to work with theReturnDataSpecialist agent to answer the inquiry.The Supervisor used one of the tools available to give the ReturnDataSpecialist the order number for the item the customer wanted to return:Customer John Smith wants to return the popcorn bowl from order BAJ1UHPC5W.4. ReturnDataSpecialist responseAlong with the Supervisor's final response, you can see every response formed by other agents on the team. In this example, the ReturnDataSpecialist used the order number it received from the Supervisor to form a response to the customer.Understand guardrailsBy setting guardrails, you can prevent your AI agents from mentioning or discussing competitors in customer interactions. In addition to protecting against mentions of competitors, these guardrails work alongside standard safeguards within the product to ensure AI agents don’t disclose sensitive information. You can see if any guardrails were detected in the trace for each message.The following guardrails are available for AI Agents.GuardrailDefinitionCompetitorDetects only explicit mentions of the companies defined in the Competitor Guidance field for the team's settings. It's fine to mention other company names in a neutral or positive way.ToxicityDetects toxic messages including profanities, obscenities, threats, insults, and identity attacks, including examples of obfuscated profanities using alphanumeric and non-alphanumeric characters.NSFWDetects content that is not safe for work and non-inclusive, including sexual comments or offensive notes around age, gender, or other categories and protected classes.SecretsDetects proprietary information used to access confidential systems like passwords, private keys, or other credentials. This guardrail does not flag messages related to verification processes, such as "six-digit passcode", credit card numbers, email addresses, phone numbers, or other personal information.

AI Agents are designed to automate conversations and improve customer interactions. They efficiently handle tasks and make decisions with a human-like approach, leading to quicker solutions and more satisfied customers.The AI Agent report helps you understand the quality of responses generated by AI Agents and view key metrics such as the number of conversations being handled by AI Agents and through which channels.Who can access this feature?User typesManagers can access reports.Learn more about AI Agents in this article.In this article:Overview metricsMessage Activity chartConversations by channel chartActive AI Agents chartConversation Activity chartTo access the AI Agent report, go to Reporting and then select AI Agent.Overview metricsThe following metrics are an overview of AI Agent interactions.The following metrics are available:The number of customers that interacted with an AI Agent.The total number of conversations in which an AI Agent was involved.The total number of messages that an AI Agent processed.Message Activity chartThis chart shows the message activity by channel and the average time between an inbound message and the AI Agent’s reply to that message.Conversations by channel chartThis chart shows the breakdown of conversations handled by an AI Agent.Active AI Agents chartThis chart breaks down AI Agents by execution volume.Note: An AI Agent can execute multiple times for a single event.Conversation Activity chartThis chart shows the activity of AI Agent-related conversations by channel.

AI Agents can connect to external systems using tools powered by OpenAPI specifications, allowing them to retrieve real-time data and provide more accurate responses to customers. By integrating an API with an OpenAPI spec, you can extend your AI Agent’s capabilities and perform actions such as:Get customer data from external systemsPlace orders on external e-commerce systemsAdjust bookings on travel booking systemsCheck stock levels from warehouse inventory systemsWho can access this feature?User typesAdmins can access the AI Agents Studio page.This article provides an example specification you can use as a guideline to create your own.In this article:PrerequisitesExample OpenAPI specificationTune request parametersUnsupported OpenAPI fieldsCreate a tool that connects to the OpenAPI specPrerequisitesBefore creating the tool, you must have an OpenAPI specification for the API you want to use with AI Agents. The specification must adhere to the OpenAPI format.Example OpenAPI specificationThis example uses the free OpenWeather API to get current weather data for a particular city. When using a tool with this specification, the AI Agent will determine which city to get weather for by reasoning about the conversation and the user's intent. For example, let's say the following message is sent to an AI Agent that has this tool assigned: "Hi, what's the weather in London?"The AI Agent will:Determine that the user intends to obtain current weather information about LondonLook at which tools it has to help the customerIdentify that it has a tool assigned to itDetermine that this tool will address the customer’s need for current weather in LondonCall the tool and sets the input parameter of q to “london”The tool makes a GET request to the url: https://api.openweathermap.org/data/2.5/weather?q=london&lang=en. For more information on parameters and tuning them to your needs, see the section below.Return the weather data output from the tool in a friendly tone to the customerExample OpenWeather API:openapi: 3.0.0 info: title: OpenWeatherMap API description: Sample OpenWeather API. version: "2.5" servers: - url: https://api.openweathermap.org/data/2.5 paths: /weather: get: summary: Call current weather data for one location description: Get the current weather info operationId: CurrentWeatherData parameters: - $ref: "#/components/parameters/q" - $ref: "#/components/parameters/lang" responses: "200": description: Successful response content: application/json: schema: $ref: "#/components/schemas/200" components: parameters: q: name: q in: query description: For the query value, type the city name and optionally the country code divided by comma; use ISO 3166 country codes. required: true schema: type: string lang: name: lang in: query description: the language to return the city name and description in when retreiving the weather required: true schema: type: string enum: - en schemas: "200": title: Successful response type: object properties: coord: $ref: "#/components/schemas/Coord" weather: type: array items: $ref: "#/components/schemas/Weather" description: (more info Weather condition codes) base: type: string description: Internal parameter example: cmc stations main: $ref: "#/components/schemas/Main" visibility: type: integer description: Visibility, meter example: 16093 wind: $ref: "#/components/schemas/Wind" clouds: $ref: "#/components/schemas/Clouds" rain: $ref: "#/components/schemas/Rain" snow: $ref: "#/components/schemas/Snow" dt: type: integer description: Time of data calculation, unix, UTC format: int32 example: 1435658272 sys: $ref: "#/components/schemas/Sys" id: type: integer description: City ID format: int32 example: 2172797 name: type: string example: Cairns cod: type: integer description: Internal parameter format: int32 example: 200 Coord: title: Coord type: object properties: lon: type: number description: City geo location, longitude example: 72.85 lat: type: number description: City geo location, latitude example: 19.01 Weather: title: Weather type: object properties: id: type: integer description: Weather condition id format: int32 example: 803 main: type: string description: Group of weather parameters (Rain, Snow, Extreme etc.) example: Clouds description: type: string description: Weather condition within the group example: broken clouds icon: type: string description: Weather icon id example: 50d Main: title: Main type: object properties: temp: type: number description: "Temperature. Unit Default: Kelvin, Metric: Celsius, Imperial: Fahrenheit." example: 293.25 pressure: type: integer description: Atmospheric pressure (on the sea level, if there is no sea_level or grnd_level data), hPa format: int32 example: 1019 humidity: type: integer description: Humidity, % format: int32 example: 83 temp_min: type: number description: "Minimum temperature at the moment. This is deviation from current temp that is possible for large cities and megalopolises geographically expanded (use these parameter optionally). Unit Default: Kelvin, Metric: Celsius, Imperial: Fahrenheit." example: 289.82 temp_max: type: number description: "Maximum temperature at the moment. This is deviation from current temp that is possible for large cities and megalopolises geographically expanded (use these parameter optionally). Unit Default: Kelvin, Metric: Celsius, Imperial: Fahrenheit." example: 295.37 sea_level: type: number description: Atmospheric pressure on the sea level, hPa example: 984 grnd_level: type: number description: Atmospheric pressure on the ground level, hPa example: 990 Wind: title: Wind type: object properties: speed: type: number description: "Wind speed. Unit Default: meter/sec, Metric: meter/sec, Imperial: miles/hour." example: 5.1 deg: type: integer description: Wind direction, degrees (meteorological) format: int32 example: 150 Clouds: title: Clouds type: object properties: all: type: integer description: Cloudiness, % format: int32 example: 75 Rain: title: Rain type: object properties: 3h: type: integer description: Rain volume for the last 3 hours format: int32 example: 3 Snow: title: Snow type: object properties: 3h: type: number description: Snow volume for the last 3 hours example: 6 Sys: title: Sys type: object properties: type: type: integer description: Internal parameter format: int32 example: 1 id: type: integer description: Internal parameter format: int32 example: 8166 message: type: number description: Internal parameter example: 0.0166 country: type: string description: Country code (GB, JP etc.) example: AU sunrise: type: integer description: Sunrise time, unix, UTC format: int32 example: 1435610796 sunset: type: integer description: Sunset time, unix, UTC format: int32 example: 1435650870Tune request parametersAI Agents can provide different arguments for tools for different use cases. In this example spec, the AI Agent determines that it will need to call the GetWeather tool with the q parameter set to “london” to address the customer’s needs. There are two ways to set up different input parameters when configuring any tool in the AI Agent Studio:Reasoning attributesReasoning attributes are populated by the AI Agents when the tool is used to help a customer, based on all of the information at its disposal, such as:Conversation data (messages, etc)Customer dataKnowledge base articlesPrior tool outputPreset AttributesPreset attributes are static, and do not change when the tool is used. When an AI Agent uses a tool with preset attributes, the tool is always called with the preset value configured for the preset attribute. A common use case for this is when you want the tool to always use the same value.When it comes to OpenAPI tools that send requests to other systems, these parameters can be used to form a query string, a request body, a path parameter, or a header value. In the example above, we used a mixture of Preset and Reasoning attributes to fill in the query string of the request, with the primary goal of retrieving weather information for any city, in English.Unsupported OpenAPI fieldsNot all fields in an OpenAPI specification are pertinent when considering tools to be used by an AI Agent. Kustomer has simplified this process by supporting only the most critical fields from the expansive OpenAPI Specification. The below fields and features are not supported today:security and securitySchemasKustomer AI Agent Studio allows you to create an Authentication through UI form, therefore we do not need to provide Security information in the OpenAPI Specification itselfrequestBodiesSince AI Agent Studio creates one tool from your OpenAPI Specification, there is no need to use this field for sharing amongst different OpenAPI pathsCookie parametersOptional fieldsAI Agents work best when they have explicit instructions and parameters to work from. We find it most effective to create tools that have required parameters. If you need different variations of input parameters, we suggest creating multiple tools. This means every parameter needs to be marked as “required” in your OpenAPI Specification.Multiple ServersTools should only talk to one API, therefore, we only support one Server URL.Multiple Paths or OperationsAs mentioned previously, AI Agent Studio only supports one tool at a time when created from an OpenAPI Specification. Tools have a 1:1 relationship with Paths and Operations in an OpenAPI Specification.Create a tool that connects to the OpenAPI specOnce you have your OpenAPI specification, you can create a tool in AI Agent Studio that uses it to respond to customer inquiries.Go to Settings > Kustomer IQ > AI Agent Studio.Select the Tools tab and go to OpenAPI.Enter a name and description for the tool. These should be descriptive of what the tool will do, as both of these fields will be used by your AI Agent to determine when and how to use the tool. Since this example tool will retrieve weather data, we will name it “Get Weather Data” and set the description to “This tool is used to get current weather data for a particular city”.Next, select the authentication required to connect to your API by selecting Add Authentication, and setting up the required fields. Currently, the supported authentication methods are:Bearer TokenAPI KeyBasic AuthenticationFinally, enter the API spec in the area provided. Select Create. 

In addition to your public knowledge base, your AI Agent teams can use public URLs to respond to customer inquiries. Who can access this feature?User typesAdmins can access the AI Agent Studio page.Learn more about AI Agents in this article. To learn how to use the data source when creating an AI Agent team, see this tutorial.In this articleAdd data sourcesManage your data sourcesFAQAdd data sourcesThe first step is creating a team to respond to these inquiries. Go to Settings   > KIQ > AI Agent Studio and select the Data Sources tab.Enter a name and brief description of the types of content the external URLs have.Next, enter the public URL(s) you want to use by selecting Add Url as https://www.urlhere.com.Note: Some public web pages automatically block other sites to prevent web crawlers and search engines from indexing their contents. These pages cannot be used as a data source. Select Create.Manage your data sourcesOnce you add a new data source, you can view its status from the Data Sources tab. Sources that are Indexed are ready for AI Agents to use to respond to inquiries.Note: Refresh your browser to update the sync status.If you want to add additional URLs to an existing data source, select Options  > Edit. Once you submit your changes, sync them by selecting Refresh .If you want to delete an existing data source, select Options  > Delete. FAQQ. How long does it take to sync a new data source?In most cases, a new data source is synced within a minute. To confirm it was synced, refresh your browser.Q. Can I add a sitemap as a URL?No, we do not support sitemaps at this time. You must enter URLs individually to upload them as a data source.Q. Are changes to the content on a page indexed automatically in Kustomer?No, you need to Refresh  the data source that contains that URL to ensure the latest content is used by AI Agents to respond to clients.Q. My page has an infographic, will AI Agents use that to respond to a customer?AI Agents only use text found on a page to respond. Any text found in videos, images, or descriptive alt text will not be used.Q. I embedded a Google sheet on one of my pages, can AI Agents read the embedded sheet?No, content that is embedded into a page cannot be used to respond.

You can transfer conversations to an AI Agent team in two ways: manually by assigning the AI Agent team or via an automated workflow. These options let you decide how and when AI Agents get involved in customer interactions.Who can access this feature?User typesAdmins for organizations that have the AI Agents for Customers add-on can access these features.In this articleManually assign  a conversation to an AI Agent teamUse Customer Assist to hand off a conversation to an AI Agent teamUse a workflow to hand off a conversation to an AI Agent teamManually assign a conversation to an AI Agent teamYou can turn on a setting allowing agents to manually assign a conversation to an AI Agent team.Note: The team will only be visible on conversations that meet the deployment conditions for the AI Agent team. For example, if your team is set only to respond to email conversations, you will only see the AI Agent team available in the Assign menu during an email conversation.Go to Settings  > KIQ > AI Agent Studio.Create a new AI Agent team or Edit an existing one.Select the Settings tab and turn on Enable human-to-AI Agent Team handoff.Select Submit.Your AI Agent team will now be an option in a conversation's Assign to a User drop-down menu.Once you pick the team, you can leave a note for the AI Agent to give it more context as to why it's being assigned the conversation. For example, if the inquiry was about a return, you can leave a note telling the AI Agent that they can tell the customer their refund was approved. The note will also appear in the conversation.If the human to AI Agent handoff setting is turned on, you can also @ mention an AI Agent team in any note within a conversation. When the conversation is handed over, the AI Agent will reference all notes where it was mentioned to gather relevant information and respond to the customer. Use Customer Assist to hand off a conversation to an AI Agent teamYou can use the Transfer to AI Agent team interaction to transfer a conversation to an AI Agent team.To learn more about this interaction, see Use Customer Assist to hand off a conversation to an AI Agent team.Use a workflow to hand off a conversation to an AI Agent teamYou can use the AI Agent Handoff action to transfer a conversation to an AI Agent team.For a tutorial on how you could use this action, see Tutorial: Use a workflow to hand off a conversation to an AI Agent team.

You can use workflows to transfer conversations to an AI Agent team automatically. In this tutorial, you'll learn how to use the Handoff to AI Agent workflow action to automate this process.Who can access this feature?User typesAdmins can access the Workflows page. This workflow action is available with the AI Agent for Customers add-on.In this articlePrerequisitesExample workflowPrerequisitesTo use this action to hand off a conversation, you must have an AI Agent team set up and deployed. You should also be familiar with workflows.Example workflowThis example will walk you through building a workflow that will transfer conversations about order status to an AI Agent team.Go to Settings > Platform > Workflows and select Add Workflow.Select Custom Workflow, enter a name, and select Create.For the first Trigger step, select Kustomer as the Trigger App and Message Created as the Trigger Event.Add a Condition Step for Direction Type Equals Initial In. For this example, we want only conversations where customers are inquiring about their order status to be transferred. We can do that by adding another Condition Step and outlining keywords in the subject such as track order or status.Next, add an Action step with Kustomer as the Action App and Conversation: Update as the Action Event. Select the tag you want to apply from the Tags field. For this example, we want the workflow to apply a WISMO tag.Finally, add an Action step with Kustomer as the Action App and AI Agent Handoff as the Action Event.Select the AI Agent team to which you want to transfer the conversation from the User drop-down menu.

If you're already using a conversational assistant, you can incorporate AI Agent teams into your existing setup to provide more intelligent, dynamic responses. This feature is especially useful for transitioning from a deterministic conversational assistant to a more advanced, reasoning-based AI Agent. In this article, you'll learn how to access the Transfer to AI Agent interaction and relevant information to keep in mind when incorporating this interaction into your assistant.Who can access this feature?User typesAdmins can access the Customer Assist page. This interaction is available with the AI Agent for Customers add-on.In this articlePrerequisitesUnderstand the transfer interactionExample conversational assistantUnderstand rule orderPrerequisitesTo use this interaction to hand off a conversation, you must have an AI Agent team set up and deployed. You should also have an existing conversational assistant built and understand conversational assistant rules.Understand the transfer interactionThe Transfer to AI Agent interaction is available for the chat, SMS, WhatsApp, and Kustomer Voice channels on both Lite and Premium assistants.You can find it when building an assistant by selecting Add End Interaction. Your deployed AI Agent teams will appear on the list.Note: To transfer conversations successfully, the AI Agent team must be deployed for the same channel as the assistant. However, the AI Agent will only respond if its own deployment conditions are met at the time of transfer. For example, if the AI Agent is set to reply only in chat and only for VIP customers, both conditions must still hold true when the conversation is handed off.Example conversational assistantImagine you have a conversational assistant set up that asks customers "How can I help" and presents them with a list of the most frequently asked topics as buttons, with one of those options being "Order status". You also have an AI Agent team that handles these types of inquiries. By transferring these types of inquiries to your AI Agent team, you can enhance the customer experience. The AI Agent can pull real-time order details and provide a personalized update, improving efficiency and making responses more helpful.Understand rule orderThe order of your rules determines whether an AI Agent team or a conversational assistant responds to a conversation first.When you deploy an AI Agent team, a rule is automatically created to ensure the team engages with conversations first. You can find this rule by going to Settings > Kustomer IQ > Customer Assist and selecting the Rules tab. In our example above, we want the conversational assistant to respond to a customer first and present the list of topic options to choose from. To ensure this happens, the conversational assistant's rule must be ranked higher than the AI Agent team's rule.

Tools refer to specific features or data sources that AI Agents can use to assist with customer requests. Think of tools as the AI’s way of reaching into your organization’s resources to find the right answer—whether it’s pulling from your knowledge base articles or checking customer data from the platform. These tools allow the AI to access information found in your organization, helping it respond with accurate and helpful answers.Who can access this feature?User typesAdmins can create custom tools for AI Agent Teams.In this articleTypes of available toolsAvailable fieldsConversation fieldsCustomer fieldsKObject fieldsSupported channelsTypes of available toolsTools enable AI Agents to give more specific, relevant, and helpful responses to customer inquiries. Without tools, it would be limited to responding to customers using generic or basic information. AI Agents can handle more complex requests by accessing your organization's data (like purchase history or a help article), saving your agents time and providing a better customer experience.AI Agents have access to the following tools in your organization:Standard objectsThese are the core data objects in Kustomer, like customer names and conversation tags. AI Agents can access Conversation and Customer data to personalize its responses or offer more context. For example, it might use a customer's name for a more personalized response.Custom objects, or KObjectsEvery business has unique data needs, and custom objects let you track data that doesn’t neatly fit into standard categories. AI Agents can use KObjects in the same way they use standard objects, providing highly tailored responses based on your organization’s specific setup. For example, it can find a customer's latest purchase information from an Order object Custom attributesCustom attributes are additional fields you’ve added to your data model, and the AI Agent can reference them to provide more detailed or accurate responses. For instance, if you’ve added a custom attribute for a customer’s loyalty status, the AI Agent can include that in its replies.Available fieldsAI Agents can access various fields in your standard and custom objects and perform various actions on different fields, such as retrieving, updating, or creating information. However, not all actions are supported for every field. For instance, while the AI Agent can retrieve (Get) a customer’s email address, it may not be able to update this field directly. Below you’ll find a detailed breakdown of the fields that support each type of action.FieldUpdateConversationAssigned TeamXAssigned UserXBrandXNameXPriorityXQueueXTagsXFieldUpdateGetCustomerBirthdayAtXXDefault LangXDisplay NameXXEmailsXExternal IdsXFacebook IdsXFirst NameXXGenderXXInstagram IdsXLast NameXXLocationsXNameXXPhonesXShared EmailsXShared PhonesXShared SocialsXSocialsXUrlsXUsernameXWhatsAppsXFieldGetUpdateCreateKObjectAssigned UserXDescriptionXXXStatusXTitleXXXCustom attributesXXXSupported channelsAI Agents offer support across different communication channels. However, not all channels support the same operations on customer data or custom objects. Depending on the channel used, the AI Agent may be able to perform specific actions, such as retrieving a customer’s name or updating certain fields, while other operations may be restricted. Access to these tools also depends on whether the customer is verified. While most channels like email and SMS can be accessed without verification, certain channels—such as Kustomer Voice—require verification to ensure secure usage of the tool.The table below outlines which fields and operations are available for each channel, as well as the level of verification necessary.KlassOperationEmailSMSWhatsAppAuthenticated chatAnonymous chatKustomer VoiceCompanyGetNo verificationNo verificationNo verificationNo verificationVerificationVerificationUpdateVerificationVerificationNo verificationNo verificationVerificationVerificationCreateNot supportedNot supportedNo verificationNo verificationNot supportedNot supportedCustomerGetNo verificationNo verificationNo verificationNo verificationVerificationVerificationUpdateVerificationVerificationNo verificationNo verificationVerificationVerificationConversationGetNo verificationNo verificationNo verificationNo verificationVerificationVerificationUpdateNo verificationNo verificationNo verificationNo verificationNo verificationNo verificationKobjectGetNo verificationNo verificationNo verificationNo verificationVerificationVerificationUpdateVerificationVerificationNo verificationNo verificationVerificationVerificationCreateVerificationVerificationNo verificationNo verificationVerificationVerification