Enjo AI for IT teams

How to configure and use Enjo's AI Agent to handle IT service desk requests end-to-end, from automated resolution to escalation and reporting.

IT teams handling high volumes of service requests can use Enjo's AI Agent to resolve common requests end-to-end, route exceptions to human agents with full context, and surface patterns in request data.

Key Capabilities

Answer employee queries

• Searches connected knowledge sources including Confluence, Notion, SharePoint, uploaded documentation, and past tickets.

• Generates responses with citations grounded in your approved knowledge

• Responds directly in Slack, Teams, or web without creating a ticket when a match is found

• Responses can be tuned using feedback signals and training examples via the Training module

Modules: Knowledge, Answers, Training

💡Tip: Connect Confluence or SharePoint as your primary knowledge source before going live. The AI Agent's response accuracy depends on how well your knowledge base covers common IT requests. Use the Knowledge module to review what is indexed and fill gaps.

Route and escalate requests

• Determines when and where to hand off using rules configured in the Escalation module

• Creates a ticket in JSM, ServiceNow, or another connected ITSM on escalation

• Passes the full conversation context to the assigned agent so they start with complete history

• Routing logic can be extended with multi-step AI Flows for more complex scenarios

Modules: Escalation, AI Flows

💡Tip: Set up at least one fallback escalation route before enabling the AI Agent in production. If no escalation rule matches, requests may go unhandled. The Escalation module lets you define a catch-all rule to route unmatched requests to a default queue.

Capture IT requests in the ticketing system

• Classifies incoming requests and creates tickets in your connected ITSM (Jira Service Management, ServiceNow, or Freshservice)

• Populates ticket fields directly from the conversation

• Sends employees status updates as the ticket progresses

• The Workflows module handles triage, tagging, and follow-up steps

Modules: AI Actions, Workflows

💡Tip: Run Desk Assessment before configuring ticket classification. It analyzes your existing ticket history to identify the highest-volume request types, so you can prioritize which categories to configure first.

Surface patterns and gaps

• Global Insights shows trends in topics and sentiment across all requests

• Performance Insights tracks AI and human resolution accuracy

• Executive Insights surfaces ROI metrics including time saved and deflection rate

• All Insights modules are available in the Insights product, separate from the AI Agent

Modules: Global Insights, Performance Insights, Executive Insights

💡Tip: Check Performance Insights weekly during your first month. It will surface which request types the AI Agent is handling well and where accuracy is low, helping you prioritize improvements in the Training module.

Note: Escalated requests land in the Inbox, which is Enjo's shared workspace for human agents. See Inbox for workload management, SLAs, and views. For in-tool AI assistance within JSM or Zendesk, see Agent Assist.

Supported channels

The AI Agent supports Slack, Microsoft Teams, and web (Enjo widget) for IT deployments. Slack is the primary channel with the full feature set including slash commands, emoji actions, and approval flows. Teams supports the same core workflows; some Slack-native features may not be available.

💡Tip: If your organization uses both Slack and Teams, start with Slack. It supports the full feature set including emoji actions and approval flows. Add Teams as a second channel once your core workflows are stable.

Supported integrations

Common IT integrations include Jira Service Management (primary ITSM for most deployments), ServiceNow, Freshservice, Okta (account unlock, access provisioning), and Microsoft Azure AD / Entra ID. Knowledge sources include Confluence, SharePoint, Google Drive, and Guru. See Integrations for the full list and setup guides.

💡Tip: Configure Okta and Azure AD as System Actions, not knowledge sources. These integrations allow the AI Agent to take real-time actions such as unlocking accounts or provisioning access during a conversation, which is distinct from indexing content for answers.

What's next ?

Once you understand the core capabilities, the typical setup path for an IT deployment is:

1. Connect your knowledge sources: Index Confluence, SharePoint, or uploaded docs to ground AI responses.

2. Configure escalation rules: Define when the AI Agent hands off and which ITSM receives the ticket.

3. Enable channels: Connect Slack or Teams so employees can reach the AI Agent.