> For the complete documentation index, see [llms.txt](https://docs.siit.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.siit.io/ai/triage-agent.md). # Triage Agent The Triage Agent handles request intake conversationally in Slack, Teams and the employee Portal. It deflects with a knowledge article when one exists, or creates a complete, structured request The Triage Agent handles request intake conversationally. It sits in your Slack and Teams channels and in the employee Portal, understands what an employee needs, deflects with a knowledge article when one exists, and otherwise creates a complete, structured request: right service, form fields filled, priority set, tags applied. {% embed url="" %} ### What you get * **Knowledge deflection that reads the conversation.** The agent checks your knowledge base first and suggests the relevant article with a summary. Employees accept or decline in plain language, no exact phrasing or button clicks required. If the article resolves the problem, no request is created. * **Intelligent service matching.** The agent reads your Service Catalog to route each request to the right service. * **Complete form compliance.** Every required field, including custom and multi-select fields, is collected through conversation and validated against your form's constraints before the request is created. * **Priority and tags, set automatically.** Derived from the conversation content and your service configuration. * **Request on behalf.** Managers can open requests for their direct reports. The agent identifies the requester, confirms, and adds the manager as a follower. * **Full conversation on the timeline.** The entire AI exchange is attached to the request, so admins have complete context without opening Slack. * **Article suggestions on the record.** When the agent offers an article, the request shows that it was suggested and whether the employee turned it down, so you can see the deflection that was attempted before the request was created. The same detail appears in the conversation summary. * **Works where employees already are.** Slack, Microsoft Teams, and the employee Portal. ### Setup * **Channels:** which Slack or Teams channels the agent listens to. * **Knowledge behavior:** whether the agent checks the knowledge base before moving to intake. * **Request behavior:** **Try to create directly** skips the confirmation step; **Always show pre-filled form** lets the employee review and submit. * **Stale reminders:** how long the agent waits before nudging an abandoned conversation. For step-by-step setup, see the [Triage Agent guide](https://help.siit.io/triage-agent-guide) in the Help Center. ### Best practices * Start with one channel. Run real requests through it, then expand once matching looks right. * Write service `About` fields in plain language, listing the request types each service covers. This is the single highest-leverage change for accuracy. * Keep forms to 3 to 6 distinct, clearly labelled fields. Multi-select options should describe distinct scenarios, not synonyms. * Keep your knowledge base current. The agent surfaces what you publish, stale articles included. * If an article you expected isn't suggested, see [An article isn't suggested](/ai/triage-agent/article-not-suggested.md). * Begin with **Always show pre-filled form** if you want employees to confirm before a request is created, then move to direct creation once you trust the matching. #### Availability The Triage Agent is part of the **Pro plan**, with a 14-day free trial available on any plan. AI provider selection, permissions, and audit behaviour are covered in the [AI trust model](https://docs.siit.io/ai/ai-trust-model). --- # 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 by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.siit.io/ai/triage-agent.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. 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.