Iris AI overview: what it is and how it works

Applies to: Iris AI. Last reviewed: November 2026.

Iris AI is the next version of Khoros Care and Khoros Marketing. It brings your community, your social channels and your customer conversations into one workspace, with an AI layer that reads, prioritizes and drafts replies for every conversation. This article explains what Iris AI is, how it works, and the terms you will see across the rest of the Iris AI Docs.

What Iris AI does

• One inbox for every channel. Posts from your Khoros Community, comments and messages on your social accounts, mentions of your brand and items from RSS feeds arrive in a single, prioritized queue.
• AI on every conversation. Each item can be summarized, tagged, scored for sentiment, emotion and relevance, and given suggested draft replies before an agent opens it.
• A human always decides. AI drafts are suggestions. An agent reviews, edits and sends every reply. AI features are off until your administrator turns them on.
• Built-in measurement. Volume, response times, SLA compliance, sentiment and team productivity are recorded automatically and shown in Analytics.

The two sides of the platform

Side	What it covers
Care (the Agent workspace)	Everything that comes in: community posts, comments, mentions, reviews and direct messages. Each conversation becomes a case that is routed to the right team and handled by an agent.
Marketing (Publisher)	Everything you send out: planning, approving and scheduling posts to your social accounts.
AI (across both)	Summaries, sentiment, emotion, language, tags, priority, relevance and draft replies, plus AI help when writing posts.

It is one platform with one sign-in. "Care" and "Marketing" are simply the two halves of the workspace.

How it works

How Iris AI works

1. Connect your Khoros Community and your social accounts.
2. Collect. Each connection collects new activity: community posts and replies, comments, mentions and messages, as each network makes them available.
3. Understand. New activity creates a case, or joins the existing conversation with that person, and the AI analyses it.
4. Route. Your tags, work queues and priority rules put the case in front of the right team, most urgent first.
5. Respond. An agent opens the case, reviews the AI context, uses or edits a draft and replies from the right account.
6. Measure. Analytics records every step so you can see what is working.

Key terms

Term	Meaning	Where you set it up
Connection	A link between Iris AI and one of your sources (a Khoros Community, an X account, an RSS feed). It collects new activity continuously.	Admin > Social Connections
Message, conversation, case	A message is one item (a post, comment or DM). A conversation is the back-and-forth with one person on one channel. A case is the work item an agent handles for that conversation, with its status, priority, tags and history.	Cases appear in the Agent view
Tag	A label on a case, applied by your rules, by AI or by an agent. Tags drive routing.	Admin > Tags
Work queue	Where routing sends a case: the queue a team works from.	Admin > Work Queues
Priority	How urgent a case is, from P0 (highest) to P5 (lowest).	Admin > Priority
Smart view	A saved filter over the cases a user can already see. It focuses work but does not route it.	Admin > Smart Views
Role	What a person may do (for example Agent, Supervisor or Account Admin).	Admin > Users

Where to go next

Follow the articles in this order to set up your instance:

1. Signing in to Iris AI and finding your way around
2. Adding your team in Iris AI: users, roles and teams
3. Connecting channels to Iris AI, then Connect your Khoros Community to Iris AI and the guide for each social channel you use
4. Routing work in Iris AI: tags, work queues, priority and smart views
5. AI features in Iris AI
6. Handling conversations in the Iris AI Agent view
7. Your first week with Iris AI

Questions are answered in Iris AI frequently asked questions. For anything specific to your account, contact your Khoros Customer Success Manager.