Enjo Concepts
This page explains the key concepts you will encounter across Enjo. If you are setting up Enjo for the first time, read this before anything else.
Enjo is built around one core loop: an AI Agent handles incoming support requests using your knowledge, escalates what it cannot resolve to a human agent in Inbox, and improves over time as you refine its behavior. Everything else in the product supports or extends that loop.
The core loop
These five concepts are the foundation of how Enjo works. Together they describe how a request enters the system, gets resolved or escalated, and reaches a human when needed.
AI Agent
The AI Agent is the automated layer that handles incoming support requests. It searches your connected knowledge, follows the workflows you define, and takes actions in external systems when needed. When it cannot resolve a request on its own, it escalates to a human agent in Inbox with the full conversation and context attached.
You can have more than one agent in a workspace, each configured for a different team or use case. If you have multiple agents, you can set up a Router AI Agent to sit in front of them and direct incoming requests to the right one based on the content of the request.
You can also configure the agent's voice, tone, and response style to match your brand, and personalize its answers based on user attributes like role, department, or account history.
Knowledge
Knowledge is how you give the AI Agent the content it needs to answer questions. The agent only knows what you connect and approve.
You can connect existing documentation from apps like Confluence, SharePoint, and Google Drive, upload files and website URLs directly, or even add audio and video content that Enjo will extract and index. For past conversations and tickets, sources like Slack, Jira, Zendesk, and ServiceNow are also supported.
Knowledge sources can be shared across multiple agents or scoped to a specific one. You can also set a priority order so the agent references certain sources first.
💡Tip: Two specialized knowledge features, Enjo Answers and Help Center, are covered in their own sections below.
Channels
Channels are the platforms through which users interact with the AI Agent. Full AI resolution is supported on Slack, Microsoft Teams, and web chat (Enjo widget). Email routes inbound messages to Inbox for human agent handling. Each channel is connected separately and can be assigned to one or more agents.
Escalation
Escalation rules define when and how the AI Agent hands a request off to a human. You control the conditions that trigger a handoff, where the request goes (Inbox or a connected ticketing system such as Jira, Zendesk, Salesforce, ServiceNow, or Freshservice), and what context is passed along. The human agent receives the full conversation history and a structured summary.
When a connected ticketing system is configured, the agent can also automatically create and update tickets from conversations based on the rules you define.
💡Tip: Escalation does not require Inbox. You can route unresolved requests directly to your ticketing system without agents ever working in Inbox.
Inbox
Inbox is the shared workspace where human agents handle escalated requests. Each conversation arrives with the full chat history, a summary of what the AI Agent did, and any linked ticket context. Agents can review, reply, take over the conversation from the AI, or assign it to a teammate.
Inbox also supports workload balancing, SLA tracking, office hours configuration, saved views for monitoring queues, and ticket status management across connected systems.
Extending the AI Agent
Once the core loop is in place, these features let you expand what the agent can do: give it exact answers, connect it to external systems, and build multi-step resolution paths.
Enjo Answers
Enjo Answers is a built-in knowledge source for handcrafted responses. Unlike connected knowledge sources where the AI composes answers from indexed content, Answers lets you define exact responses to specific questions. There are two types:
Question-based answers: Tied to a specific question and its variations. The AI uses these when it detects a strong match to the question intent.
Topic-based answers: Cover a broad range of questions within a topic, such as "return policy" or "holiday schedule." The AI falls back to these when no confident question-based match is found.
Enjo Answers is distinct from the Help Center. Answers feed the AI Agent and are not browsable by end users. The Help Center is user-facing and browsable.
AI Actions
Use AI Actions to connect the AI Agent to external systems so it can retrieve data or take actions during a live conversation. For example, the agent could look up a user account in Salesforce, create a ticket in Jira, or unlock an account in Okta. Actions are configured visually in a no-code builder and then assigned to an agent.
AI Flows
Use AI Flows when a request requires multiple steps, conditional logic, or fallback rules to resolve. A flow defines a sequence of steps in plain language: verify identity, retrieve account details, branch based on account type, escalate if unresolved. Flows can chain together multiple AI Actions into a single resolution path.
Use AI Flows when the resolution path depends on conditions that vary per user or request.
Other surfaces
These are separate products that work alongside the core loop, giving end users and human agents their own ways to interact with Enjo's knowledge and AI.
Help Center
The Help Center is a standalone, user-facing portal where employees or customers can browse support articles and get AI-powered answers. Unlike other knowledge sources that are ingested in the background, Help Center articles are explicitly authored, versioned, and published, making them auditable and controlled.
The Help Center and the AI Agent work together: the Help Center stores and governs the content, and the AI Agent retrieves and summarizes it when answering questions. A Help Center can be public or private and configured with your own domain.
Agent Assist
Agent Assist is an AI assistant embedded directly inside your existing helpdesk tool, including Jira Service Management, Zendesk, Salesforce, and ServiceNow. It helps human agents while they work on tickets by surfacing case summaries, reply suggestions, knowledge from past resolutions, real-time sentiment signals, and one-click translation, all without leaving their tool.
Test, measure, and operate
These tools help you validate your setup before going live, monitor performance once you are live, and extend your team's capabilities inside Slack and Teams.
Playground
Use Playground to simulate user queries and verify how the agent responds before anything goes live. Run it after connecting knowledge or configuring actions to confirm the agent behaves as expected.
Insights
Insights gives you a single place to monitor how your AI Agents and human agents are performing. It covers AI resolution rates, ticket analytics, customer satisfaction, team productivity, and cost savings. Use it to identify what is working, where gaps remain, and where to expand automation next. Insights includes a report builder and data export for sharing with stakeholders.
Automation Tools
Enjo includes a set of tools for in-chat collaboration and ticket management inside Slack and Teams. These let your team create and update tickets, manage approvals, collaborate on complex cases in dedicated channels, and receive real-time alerts for SLA risks or automation failures, all without leaving their chat platform. Enjo also provides tools to analyze historical ticket patterns and score closed tickets for quality.
Administration
Administration is where you manage security, access, and compliance settings for your workspace.
Guardrails let you define what AI Agents can and cannot do or say. You can set workspace-wide restrictions that apply to all agents, and layer additional rules on top for specific agents. This is how you control topic boundaries, content filters, and escalation triggers.
Beyond Guardrails, Administration also covers role-based permissions to control what each team member can access, user management, and a searchable audit log of all configuration changes made across the workspace.
Where to go next
If you are setting up Enjo for the first time, start here:
Once your agent is handling live conversations, configure escalation rules and set up Inbox for your team.