Measure and improve
Using the Inbox to act on what needs your judgment
The Inbox is where Unless surfaces the items that need a human decision. It is not a place for routine reading, but a working queue: signals the system detected, content gaps in your knowledge, and nudges about your AI maturity. The Messages view in the Inbox is where you see these surfaced signals, content gaps, and AI-maturity nudges from the agent that require your review and action.
What lands in the Inbox
The Inbox aggregates three types of items that need your judgment:
- Signals. Events the system detects, such as a usage ceiling, churn risk, or an upgrade opportunity. These are tied to your configured Signals and often lead to a notification or a procedure.
- Content gaps and conflicts. When the Engine flags conflicting content across Sources or topics where Living Knowledge is thin, these appear as Messages waiting for human judgment. You decide which Source is correct, then archive or update the other.
- AI maturity nudges. Recommendations that tell you what to do next to increase your maturity score. These focus on your priority row or column if you have set one.
Acting on messages
Each message in the Inbox represents a decision. For content conflicts, open the message, review the source chain, and decide which Source is correct. Fix the Source and the Content library reprocesses automatically. For signals, you might configure a notification or a procedure to act on the event. For maturity nudges, the recommended action is listed with an estimated impact, and you can take it directly from the message.
The Inbox is organized into tabs for better grouping of messages, so you can work through one type of item at a time. When you set a priority in the AI maturity dashboard, the Inbox messages that tell you what to do next will focus on that priority, giving you actions relevant to the row or column you selected instead of generic recommendations.
How the Inbox connects to the rest of Unless
The Inbox is not a standalone tool. It is the human checkpoint in an otherwise automated system. The Content library flags conflicts and gaps that appear in the Inbox as Messages. The AI maturity dashboard reads from your configuration, Performance, and Business impact data, and its recommendations flow into the Inbox. The Team Assistant also surfaces alerts inside your helpdesk, but those are fired automatically by Unless and are not configured on the Notifications screen.
When you act on a message, you are not just clearing a queue. You are training the system. Every accepted draft and every override feeds back into the same Living Knowledge that powers the Customer Agent, so your team’s expertise compounds into the product.
Best practices for using the Inbox
- Check the Inbox regularly. Messages represent judgment calls that the system cannot make alone. Leaving them unaddressed means the agent continues with unresolved conflicts or gaps.
- Resolve conflicts at the Source. Do not try to edit the Content library directly. Edits happen at the Source, and the library updates on every Source change.
- Use the priority setting. If you are focused on a specific Moment or maturity level, set that as your priority in the AI maturity dashboard. The Inbox will then show you only the actions that move that row or column forward.
- Track your impact. After taking a recommended action, return to the AI maturity dashboard. The matrix updates as conditions change, and you should give it about a week of production traffic to see the score move.
The Inbox is where the human loop closes. It is where you decide what the agent should know, what it should do, and where to focus next. By acting on what lands there, you keep the system aligned with your business goals and your team’s judgment.