In this FAQ and video, we detail how the logic behind the selection of sections used by ClaudIA works and the tagging functionality (Tagger) when closing a ticket.

We also explain how to audit this data directly in the Hub, so you can gain insights through ClaudIA's Dashboard to improve tagging.

Sessions Used by ClaudIA: What They Are, Why They Matter, and How to Audit

What Are “Used Sessions”?

Every time ClaudIA responds to a customer, it consults one or more sections of its knowledge base (N1 or N2) to formulate the response.

These sections are called used sessions.

In other words, they are the blocks of knowledge that the AI used to formulate the response sent to the customer.

🎯 Why Is This Important?

This information is essential to:

Understand how ClaudIA is reasoning in each response.
Audit whether it is actually using the right content.
Train and improve the AI's behavior, identifying when sessions are being used incorrectly.
Define the conversation's tag: the session with the highest weight influences the automatic tagging algorithm.
Detect when there is a transfer to N2, and if it was based on correct content.

What Has Been Launched

Now, you can:

Directly see which sessions were used in each response from ClaudIA.
Provide positive or negative feedback on the use of these sessions.
Audit the final tag of the conversation based on the sessions used.

How to View Used Sessions

Access any ticket in the Hub.
Find ClaudIA's response.
Click the ℹ️ icon to open the response details.
You will see:
- ClaudIA's reasoning (consulted term + intention).
- The used sessions that composed the response.
- An icon identifying which sessions were used (e.g., ✅ or highlighted icon).
- If there are N2 sessions, they will also be highlighted.

Practical Example

Customer's Question:

“Do you integrate with Zendesk and Intercom?”

ClaudIA's Response:

“Yes! We have native integration with Intercom. In the case of Zendesk…”

Used Sessions:

N1 - Intercom Integrations
N2 - Zendesk Integration via API

Outcome:

ClaudIA used content from two sessions.
Since one of them was N2, the conversation was automatically escalated.
The conversation tag was: Interest in hiring, as it was the most relevant session.

How to Audit Used Sessions

When reviewing a conversation, click the ℹ️ icon on any response.
See which sessions were marked as used.
Assess if they make sense based on the response.
Click 👍 or 👎 to provide your feedback on the use of the session.

Important: This only evaluates if the marked session was used correctly, not if the response was good or bad.
If the response is incorrect, proceed with the normal audit at the end of the ticket.

What Happens with This Feedback?

It serves to improve ClaudIA's behavior, adjusting the use of sessions.
It is used to calibrate the automatic tagging algorithm.
It helps our team understand if N2 sessions are being used correctly — which affects both the AI's performance and the escalation criteria.

Quick Tips

⚠️ Sometimes, ClaudIA consults a session to confirm something, even if the response does not have literal excerpts from the content. This still counts as valid use.
If you do not recognize the content of a session in the response, mark 👎 and briefly describe why.
The session feedback does not replace the conversation audit. It complements it.

What Influences the Conversation Tag?

At the end of the ticket, ClaudIA assigns a tag based on the session used with the highest score (relevance).

You can review this tag in the audit process:

Go to the end of the conversation.
Review if the assigned tag corresponds to the content and intention.
If you want to change it, select the correct tag and submit the audit.

Where Does All This Impact?

Dashboards of addressed topics
Retention and escalation indicators
ClaudIA's performance by session or subject
Quality of automatic suggestions for improvement
Revenue (in some contracts based on sessions)