> For the complete documentation index, see [llms.txt](https://docs.harmony.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.harmony.io/settings/configuring-the-service-desk-agent.md). # Configuring the Service Desk Agent The Service Desk Agent is a conversational AI assistant that handles employee requests across your organisation. It greets employees, gathers context about their issues, and routes or escalates requests to the right team - whether that is IT, HR, facilities, finance, security, or any other department you support. This article explains how to configure the Service Desk Agent so it reflects your organisation's structure, tone, and service scope. *** ### What the Service Desk Agent does The Service Desk Agent acts as the first point of contact for employees who need workplace support. It is designed to work across all departments, not just IT. When an employee starts a conversation, the agent: * Introduces itself as a multi-department service assistant covering IT, HR, facilities, finance, security, and other areas * Asks clarifying questions to understand the nature of the request * Routes the request to the appropriate desk or team * Escalates to a human support team when the issue cannot be resolved automatically Because the agent operates across your entire workplace service platform, employees only need one place to go for any kind of support request. *** ### Accessing the Service Desk Agent settings 1. Open **Settings** in your Harmony dashboard. 2. Select **Service Desk Agent** from the left-hand navigation. From this page you can configure the agent's identity, introduction message, department coverage, routing rules, and escalation behaviour. *** ### Configuring the agent identity The agent identity controls how the assistant presents itself to employees at the start of a conversation. #### Display name Enter the name employees will see when the agent responds. This can be a neutral name such as "Service Desk" or a custom name that fits your organisation's branding. #### Introduction message The introduction message is the first thing employees see when they open a conversation. It should make clear that the agent covers all departments, so employees know they can raise any kind of workplace request. A recommended starting point: > "Hi, I'm your workplace service assistant. I can help you with requests across IT, HR, facilities, finance, security, and more. What can I help you with today?" You can edit this message to match your organisation's tone. Keep the list of departments accurate so employees are not directed to the agent for services it does not cover. *** ### Configuring department coverage The agent supports requests across multiple departments out of the box. You can control which departments are active and how each one is labelled. #### Enabling or disabling departments Use the department toggles to turn coverage on or off for each area. Departments available by default include: * IT * HR * Facilities * Finance * Security If your organisation uses different naming conventions - for example "People Operations" instead of "HR" - you can rename each department using the **Label** field next to the toggle. The agent will use your custom labels when introducing itself and when confirming which team it is routing a request to. #### Adding custom departments To add a department that is not listed by default: 1. Scroll to the bottom of the department list and select **Add department**. 2. Enter a name for the department. 3. Configure its routing destination (see Routing rules below). 4. Save your changes. *** ### Configuring routing rules Routing rules determine which team or queue receives a request after the agent has gathered enough context to categorise it. #### How routing works The agent uses the content of the conversation to match the request to a department. Once a match is made, it confirms the routing with the employee using language such as: > "I'll send this to the Facilities team now. They'll be in touch shortly." The routing confirmation messages reference your configured department labels automatically, reinforcing that support is available across the whole workplace service platform. #### Setting a routing destination For each active department, you can specify: * **Queue or team name** - the internal queue or team the request is sent to * **Routing confirmation message** - the message shown to the employee when their request is routed (optional, overrides the default) To edit a routing destination: 1. Select the department from the list. 2. Under **Routing**, enter the target queue or team name. 3. Optionally edit the routing confirmation message. 4. Save your changes. *** ### Configuring escalation behaviour When the agent cannot resolve a request or the employee asks to speak to a person, the agent escalates the conversation to a human agent. Escalation messages are written to refer to the relevant support team rather than a specific department, so they remain accurate regardless of which department is handling the request. #### Default escalation message The default message employees see when a conversation is escalated is: > "I'm connecting you with a member of the support team now. Please hold on." You can customise this message in the **Escalation** section of the agent settings. Avoid replacing "support team" with a single department name unless the agent is scoped to that department only, as it may cause confusion for employees raising requests in other areas. #### Escalation triggers You can configure the conditions that trigger an escalation. Available triggers include: * Employee explicitly requests a human * Agent confidence falls below a set threshold * Request type is marked as always-escalate for a specific department * Conversation reaches a maximum number of turns without resolution To adjust triggers, select **Escalation triggers** and enable or disable each option as needed. *** ### Testing the agent configuration Before making changes live, use the built-in preview tool to test how the agent behaves. 1. In the Service Desk Agent settings, select **Preview**. 2. Type a sample request in the conversation window. 3. Confirm that the agent introduces itself correctly, routes to the expected department, and uses the right escalation language. Test at least one request per active department to make sure routing and messaging are consistent across all areas. *** ### Saving and publishing changes Changes to the Service Desk Agent configuration are saved as a draft until you publish them. This means you can make multiple edits before any of them go live. * Select **Save draft** to store your changes without publishing. * Select **Publish** to make the current configuration live for all employees. Published changes take effect immediately. There is no additional deployment step required. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.harmony.io/settings/configuring-the-service-desk-agent.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.