Salesforce is used for various needs of different teams in an organization, such as marketing, sales, commerce, service, and IT work as one from anywhere using the chat module. Kore.ai XO Platform’s Salesforce Agent integration allows you to seamlessly hand off the virtual assistant conversations to your live agents on Salesforce.

Use the configuration steps below as general guidelines to enable the Salesforce Agent integration. You may notice minor differences based on the version of Salesforce you are using.

Overview

Kore.ai’s Virtual Assistant (VA) serves as the central point of interaction for the users. The VA facilitates communication and interaction between the user and the Salesforce agent using the chat module. In addition to its regular capabilities, the VA acts as an intermediary between the users and the Salesforce agents to hand over conversations seamlessly. The VA also allows you to integrate with the Salesforce platform.

Configuration

The XO Platform supports the following Salesforce integration methods for agent transfer:

• Live Chat (Legacy): This method enables agent transfer using Salesforce’s traditional live chat interface. It suits users with existing Live Chat setups and offers real-time messaging between virtual assistants and human agents.
• Messaging for In-App and Web (MIAW): This is Salesforce’s latest messaging service, built for seamless and persistent conversations across in-app and web channels. MIAW offers an enhanced customer experience and is the default option for new integrations.

Before configuring a Salesforce agent, users must be enabled with the Service Cloud User license to ensure all options are displayed in the Customize section of the left navigation.

Configure Live Chat

To configure the Salesforce agent, follow these steps:

Step1: Set Up a Salesforce Live Agent

Set up a Salesforce Live Agent to receive messages.

1. Log in to your Salesforce account.
2. From the main menu on the top, navigate to Setup > Customize > Chat Agent > Settings.
   Note: Make sure you switch to the Classic view; the Live Agent settings do not come up under the Lightning Experience view. For detailed instructions on setting up Live Agent, visit:https://help.salesforce.com/s/articleView?language=en_US&id=live_agent_intro.htm
3. Find the Live Agent section under Build > Customize > Chat > Chat Agent Configurations.
   
4. To add a new Live Agent configuration, click the New button.
5. In the New Live Agent Configuration form, you can enter the details for the new configuration, such as the name and description.
6. Once you have entered the required details, click the Save button to save the new Live Agent configuration.

Step 2: Retrieve the Live Agent URL/API Endpoint

To retrieve the live agent URL/API endpoint, follow the steps:

1. From the main menu on the top, navigate to Setup > Customize > Chats > Chat Sessions > Deployments.
   
2. In the Chat Deployments page, you will see a list of all the configured Live Agent deployments.
3. To get the Live Agent URL/API endpoint, click the name of the Live Agent deployment that you want to use.
4. In the Deployment Code region, you will find the Live Agent URL/API endpoint in the Code field as shown below:
   

Step 3: Retrieve the Organization ID

To retrieve the Organization ID, follow these steps:

1. From the main menu on the top, navigate to Administer > Company Settings > Company Information.
   
2. In the Company Information page, you will see a list of details about your organization.
3. To retrieve the Organization ID, search for the Salesforce.com Organization ID field.
   
   Note: The Organization ID is a unique identifier for your Salesforce organization.
4. Copy and keep it for future use.

Step 4: Retrieve the Deployment ID

To retrieve the Deployment ID, follow these steps:

1. From the main menu on the top, navigate to Setup > Customize > Chats > Chat Sessions > Deployments.
   
2. In the Chat Deployments page, you will see a list of all the configured Live Agent deployments.
3. In the Deployment Code region, you will find the Deployment ID in the Code field as shown below:
   
4. Copy and keep it for future use.

Step 5: Retrieve the Button ID

To retrieve the Button ID, follow these steps:

1. Go to Customize > Chats Chat Sessions > Chat Buttons & Invitations page, and click the button that you want to use for this integration.
2. In the Chat Deployments page, you will see a list of all the configured Live Agent deployments.
3. In the Deployment Code region, you will find the Deployment ID in the Code field as shown below:
   
   The Button ID is the string of characters after the last slash / in the URL. For example: https://meya-dev-ed.my.salesforce.com/5731I000000PL9G
4. Copy and save it for future use.

Step 6: Enable the Salesforce Agent

1. Log in to the Kore.ai XO Platform https://bots.kore.ai/botbuilder.
2. Select a Virtual Assistant for which you want to configure the Salesforce agent.
3. Navigate to Build > Integrations > Agent Transfer and click the Salesforce.
   
4. Click the Configurations tab and enter the following credentials for the Salesforce Agent:
   • Live Agent URL: The live agent URL endpoint for the Salesforce, copied before.
   • Organization ID: The Organization ID for the Salesforce, copied before.
   • Deployment ID: The Deployment ID for the Salesforce, copied before.
   • Button ID: The Button ID for the Salesforce, copied before.

   

5. If you want to make Salesforce the default agent, enable the Default Agent System for all bot channels option.
   Note: This option is available only when multiple agents are configured for your VA.
6. Click Manage to map the channels you want to use Salesforce agent for handing over the conversations. To learn more about channel enablement, see Adding channels to your bot.
   
7. After mapping channels, click Done.
8. Click Save to enable the Salesforce agent.
9. Publish the VA. For more information, see Publishing Bot.

Execution

This section helps you verify how the conversations from the Kore.ai XO Platform are being handed over to the Salesforce agent after enabling it.

To verify the data exchange between Kore.ai’s VA and Salesforce agent, follow these steps:

1. Go to the Salesforce Chat Console and launch the Live Agent that you created in Step 1.
   Note: The Salesforce Chat Console URL is unique for your region and company. Contact your Salesforce System Administrator to find your Chat Console URL.
2. In the Salesforce Classic Chat view, click the Apps icon and select the App Launcher option.
   
3. In the App Launcher page, click the Live Agent to open the chat console as shown below:
4. Open a virtual assistant for which you have enabled the Salesforce agent.
5. Create a new Agent Transfer Node task. For example, create a Talk to Live Agent task with the LiveChats node and configure it to use the Salesforce agent. For more information, see Agent Transfer Node.
   
6. Click the Talk to Bot icon to launch the VA and enter Help.
7. Once you invoke the agent transfer dialog, the Salesforce Chat Console receives a notification alert.
8. When a user initiates the chat from the VA, it transfers the conversation to the Salesforce agent and sends a message to the user. Agent receives a notification about this on Salesforce.
9. Now, the connection between the Salesforce agent and the user has been established, and the conversation continues.
   

Configure Messaging for In-App and Web (MIAW)

To configure the Salesforce agent, follow these steps:

Step 1: Publish a Deployment

Create a custom client deployment that can be referenced in public REST API endpoints.

1. Complete the steps in Prepare a Salesforce Org for Messaging for In-App and Web:
   1. Configure Omni-Channel.
   2. Create an Omni-Channel flow to route messages.
   3. Create a Messaging for In-App and Web channels on the Messaging Settings page.
   4. Configure the Service Console to support messaging.
2. From Setup, in the Quick Find box, enter Embedded Service Deployments and select Embedded Service Deployments.
3. Click New Deployment.
4. Select Messaging for In-App and Web, and then click Next.
5. Click Custom Client, and then click Next.
6. Name your Embedded Service deployment, and customize the API name if desired.
7. Select the Messaging for In-App and Web channels that you created earlier. If you haven’t created one yet, go to the Messaging Settings page and do so.
8. Save your work.
9. Create and publish your new deployment, which takes a few minutes. Don’t navigate away from the page until it’s complete.

Step 2: Retrieve Organization ID

1. Log in to your Salesforce account.
2. Click on your profile icon in the top right corner and select Setup.
3. In the Quick Find search bar on the left panel, type Company Information and select it.
   Under Company Information, locate the field Salesforce.com Organization ID.
4. Copy the Organization ID for future use.
   
   

Step 3: Retrieve the Organization Base URL

1. Log in to your Salesforce account.
2. Click on your profile icon in the top-right corner and select Setup.
3. In the Quick Find search bar on the left panel, type My Domain and select My Domain under Company Settings.
4. Locate the field Current My Domain URL.
5. Copy this URL for integration purposes.
   
   

Step 4: Retrieve Developer Name

1. Log in to your Salesforce account.
2. Click on your profile icon in the top right corner and select Setup.
3. In the Quick Find search bar on the left panel, navigate to:
4. Platform Tools → Feature Settings → Service → Messaging → Messaging Settings.
5. Locate and click on the Messaging Channel for which you want to retrieve the Developer Name.
6. The Developer Name will be displayed within the channel details.
7. Copy the Developer Name for future use.
   
   
   

Step 6: Enable the Salesforce Agent

1. Log in to the Kore.ai XO Platform https://bots.kore.ai/botbuilder.
2. Select a Virtual Assistant for which you want to configure the Salesforce agent.
3. Navigate to Build > Integrations > Agent Transfer and click the Salesforce.
4. On the Configurations tab, select Messaging for In-App and Web(MIAW) and enter the following credentials for the Salesforce Agent:
   • Organization ID: The Organization ID for Salesforce, copied before.
   • Organization Base URL: The Deployment ID for Salesforce, copied before.
   • Developer Name: The Salesforce developer’s name, copied before.

   

5. If you want to make Salesforce the default agent, enable the Default Agent System for all bot channels option.
   Note: This option is available only when multiple agents are configured for your VA.
6. Click Manage to map the channels you want to use Salesforce agent for handing over the conversations. To learn more about channel enablement, see Adding channels to your bot.
7. After mapping channels, click Done.
8. Click Save to enable the Salesforce agent.
9. Publish the VA. For more information, see Publishing Bot.

Additional Capabilities

The Salesforce Agent integration supports metadata transfer for agent context, user-controlled session termination, and customizable system messages, ensuring faster resolutions and a consistent brand experience.

User-Controlled Session End

Allow end-users to end live agent sessions using a configurable keyword set in the Script node. The platform displays this keyword during agent handoff, giving users more control over their interactions and helping reduce agent occupancy when the session is no longer needed.

For example, in the following script, “endagentchat” is the keyword displayed to the end user to end the agent session.

let miawMetaInfo = {};
miawMetaInfo.endChatShortHandCommand = "endagentchat";
context.session.BotUserSession.miawMetaInfo = miawMetaInfo;

Customizable Standard Responses

Customize system messages shown to end-users during agent handoff and live chat sessions, including wait time alerts, agent join messages, and session closure prompts. This ensures brand consistency and enhances clarity in user interactions.

{
"en":{
"lookingForAgentMessage": "Please hold while I find an available agent.",
"agentJoinedMessage": "Agent joined the conversation.",
"conversationClosedMessage": "Your session has ended, thanks for chatting with us.",
"endChatFromUserInstruction": "To end agent conversation send endagentchat"
},
"ko":{
"lookingForAgentMessage": "잠시만 기다려주세요. 대기열에 추가해드리겠습니다.",
"agentJoinedMessage": "상담원이 대화에 참여했습니다.",
"conversationClosedMessage": "대화가 종료되었습니다. 감사합니다.",
"endChatFromUserInstruction": "상담원 대화를 종료하려면 endagentchat를 보내세요."
},
"es":{
"lookingForAgentMessage": "Por favor, espere mientras busco un agente disponible.",
"agentJoinedMessage": "El agente se ha unido a la conversación.",
"conversationClosedMessage": "La conversación ha terminado. Gracias por chatear con nosotros.",
"endChatFromUserInstruction": "Para terminar la conversación con el agente, envíe endagentchat"
},
"fr":{
"lookingForAgentMessage": "Veuillez patienter pendant que je cherche un agente disponible.",
"agentJoinedMessage": "L'agent a rejoint la conversation.",
"conversationClosedMessage": "La conversation est terminée. Merci pour votre chat.",
"endChatFromUserInstruction": "Pour terminer la conversation avec l'agent, envoyez endagentchat"
},
"de":{
"lookingForAgentMessage": "Bitte warten, während ich einen verfügbaren Agenten finde.",
"agentJoinedMessage": "Der Agent hat die Konversation beigetreten.",
"conversationClosedMessage": "Die Konversation ist beendet. Vielen Dank für Ihren Chat.",
"endChatFromUserInstruction": "Um die Konversation mit dem Agenten zu beenden, senden Sie endagentchat"
},
"ja":{
"lookingForAgentMessage": "しばらくお待ちください。利用可能なエージェントを見つけています。",
"agentJoinedMessage": "エージェントが会話に参加しました。",
"conversationClosedMessage": "会話が終了しました。ありがとうございました。",
"endChatFromUserInstruction": "エージェントとの会話を終了するには endagentchat を送信してください。"
},
"pl":{
"lookingForAgentMessage": "Proszę poczekać, gdybym szukał dostępnego agenta.",
"agentJoinedMessage": "Agent dołączył do rozmowy.",
"conversationClosedMessage": "Rozmowa zakończona. Dziękujemy za czat.",
"endChatFromUserInstruction": "Aby zakończyć rozmowę z agentem, wysyłaj endagentchat"
},
"ar":{
"lookingForAgentMessage": "يرجى الإنتظار حتى أجد أحد المستخدمين المتاحين.",
"agentJoinedMessage": "تم الانضمام إلى المحادثة.",
"conversationClosedMessage": "تم إنهاء المحادثة. شكراً لك على الدردشة.",
"endChatFromUserInstruction": "لإنهاء المحادثة مع الوكيل، قم بإرسال endagentchat"
},
"tr":{
"lookingForAgentMessage": "Lütfen bekleyin, bir kullanılabilir aracı buluyorum.",
"agentJoinedMessage": "Aracın konuşmaya katıldı.",
"conversationClosedMessage": "Konuşma sona erdi. Teşşekkürler.",
"endChatFromUserInstruction": "Aracın konuşmasını bitirmek için endagentchat gönderin."
},
"it":{
"lookingForAgentMessage": "Per favore, attendi mentre cerco un agente disponibile.",
"agentJoinedMessage": "L'agente si è unito alla conversazione.",
"conversationClosedMessage": "La conversazione è terminata. Grazie per il chat.",
"endChatFromUserInstruction": "Per terminare la conversazione con l'agente, invia endagentchat"
},
"pt":{
"lookingForAgentMessage": "Por favor, espere enquanto procuro um agente disponível.",
"agentJoinedMessage": "O agente entrou na conversa.",
"conversationClosedMessage": "A conversa terminou. Obrigado pelo chat.",
"endChatFromUserInstruction": "Para terminar a conversa com o agente, envie endagentchat"
},
"zh_cn":{
"lookingForAgentMessage": "请稍等，我正在寻找可用的代理。",
"agentJoinedMessage": "代理已加入对话。",
"conversationClosedMessage": "对话结束。谢谢您的聊天。",
"endChatFromUserInstruction": "为了结束与代理的对话，请发送 endagentchat"
},
"zh_tw":{
"lookingForAgentMessage": "請稍等，我正在尋找可用的代理。",
"agentJoinedMessage": "代理已加入對話。",
"conversationClosedMessage": "對話結束。謝謝您的聊天。",
"endChatFromUserInstruction": "為了結束與代理的對話，請發送 endagentchat"
},
"ca":{
"lookingForAgentMessage": "Si us plau, espera mentre busco un agente disponible.",
"agentJoinedMessage": "L'agent ha entrat en la conversació.",
"conversationClosedMessage": "La conversació ha acabat. Gràcies pel chat.",
"endChatFromUserInstruction": "Per acabar la conversació amb l'agent, envieu endagentchat"
},
"nl":{
"lookingForAgentMessage": "Wacht even, ik zoek naar een beschikbare agent.",
"agentJoinedMessage": "De agent is de conversatie binnengekomen.",
"conversationClosedMessage": "De conversatie is afgerond. Bedankt voor het chatten.",
"endChatFromUserInstruction": "Om de conversatie met de agent te beëindigen, stuur endagentchat"
}
}

Pass Metadata or User Information to the Salesforce Agent System

The platform automatically transfers comprehensive user context to Salesforce agents during handoffs, including usernames, case IDs, and other details. You can customize which metadata gets passed by configuring the script node before agent transfer. This seamless context sharing provides agents with complete conversation visibility, eliminating the need for users to repeat information and accelerating issue resolution.

Sample Script:

let miawMetaInfo = {};
miawMetaInfo.preChatInfo = {
"Pincode": "500081",
"City": "Hyderabad",
"Country": "India"
};
context.session.BotUserSession.miawMetaInfo = miawMetaInfo;

Follow the steps to transfer the metadata and populate it on the Salesforce dashboard:

• Step 1: Configure Salesforce Agent in AI for Service
• Step 2: Update Messaging Session Object In Salesforce
• Step 3: Map Custom Parameter In Salesforce
• Step 4: Add Pre-Chat Field in Embedded Service Deployment in Salesforce
• Step 5: Create a Flow to Update Messaging Session in Salesforce

Step 1: Configure Salesforce Agent in AI for Service

Follow these steps:

1. Log in to Platform and navigate to App Settings > Integrations > Agent Transfer > Chat > Salesforce.
2. On the Configurations tab, select Messaging for In-App and Web(MIAW) and enter the following credentials for the Salesforce Agent:
   • Organization ID: The Organization ID for Salesforce, copied before.
   • Organization Base URL: The Deployment ID for Salesforce, copied before.

     Note: If your base URL is koreaiinc.my.salesforce.com, enter it as https://koreaiinc.my.salesforce-srct.com.

   • Developer Name: The developer’s name, copied before.
     
3. Click Save.
4. Go to the Dialogflow and insert a Script Node before the Agent Transfer Node.
5. Use the sample script to configure the metadata to be transferred in the key-value pair.

Step 2: Update Messaging Session Object In Salesforce

Follow these steps:

1. Log in to the Salesforce account.
2. Navigate to Object Manager > Messaging Session > Fields & Relationships.
3. Click New to create a custom field.
4. Select the Data Type and click Next.
5. Enter the Field Label, Length, and Field Name values. Click Next.
   
6. Check the Visibility and Ready-only check boxes as required. Click Next.
   
7. Click Save. Ensure to create a custom field for each field of metainfo.
   

Step 3: Map Custom Parameter In Salesforce

Follow these steps:

1. Navigate to Feature Settings > Service > Messaging > Messaging Settings.
2. Locate and click on the Messaging Channel on which you want to map the custom parameter, for example, CustomClientTest.
   
3. In the Custom Parameter, click New map parameter.
4. Enter Parameter Name, Parameter API Name, Channel Variable Name, Data Type, and Maximum Length. Click Save.
   
5. In Parameter Mappings, click New to add your custom parameter to the flow or agent task.
6. Enter the Parameter and Flow Variable Name. Click Save.
   
7. Scroll to the top and click Active.

Step 4: Add Pre-Chat Field in Embedded Service Deployment in Salesforce

Follow these steps:

1. Navigate to Feature Settings > Service > Embedded Service > Embedded Service Deployments. The Embedded Service Deployment Settings page is displayed.
2. Click the Messaging Channel and click Edit Pre-Chat.
   
3. In the Visible Pre-Chat Fields section, click Add. Select the field type to add the created field, which is available in the Hidden Pre-Chat Fields section.
   
4. Select the Field Type and Channel Variable Name. Click Save.
   
5. Once the field is added in the Visible Pre-Chat Fields section, you can optionally mark it as required. Scroll down and click Save.
6. Click Pushish in the top right corner.

Step 5: Create a Flow to Update Messaging Session in Salesforce

Follow these steps:

1. Navigate to Process Automation > Flows.
2. Locate and click the configured flow for the Messaging Channel, or create a new flow and configure it.
3. In the Toolbox, click New Resource. Select the Resource Type, API Name, Data Type, and Availability Flow. Click Done.

   Note: Ensure the flow variable matches the API name of your custom Messaging Session field. This enables seamless metadata transfer from the Platform to Salesforce, providing agents with complete context.

4. Click Update Records and set the field values for the messaging session records.
   
5. Save the updated record as a New version and then click Activate.