Skip to main content

Builder - Sessions

Updated this week

The Sessions page in the Builder provides a detailed record of every AI Employee interaction. Use it to review lead qualification, assess agent success, and export session data.

What is a session?

A session is a single interaction β€” voice call, SMS conversation, or chat β€” between a user and the AI Employee. The system records metadata, lead flags, success assessments, and revenue metrics automatically when the session closes.

Builder Sessions vs. Customer Portal

Sessions appear in both the Builder and the Customer Portal. The Builder version includes the full 26-column set, including technical fields such as Session Id, External ID, UTMs, and Actual/Expected Result. Use the Builder when you need in-depth analysis or want to override assessments. For the Customer Portal version, see Portal - Sessions.

πŸ—’οΈ NOTE

Open the Builder Sessions page for row-level review and human assessment overrides. The Customer Portal shows a simplified view of the same data.

Access the Sessions page

  1. Open the Builder for the AI Employee.

  2. Click Sessions in the left sidebar.

Session columns

Click Columns to show or hide columns using checkboxes. The table below lists all 26 available columns.

Column

Description

Is Archived

Whether the session has been archived. Archived sessions are excluded from active reporting.

Session Type

The intent category assigned to the session.

Is Lead

true when the user expressed purchase intent. See Lead identification and Lead Value calculation.

Is Opportunity

true for valid leads without refusal, unavailability, or missing contact information. See Opportunity definition.

ASR

Additional Secured Revenue for this session. See ASR formula and example.

Is Test

Test session flag. Auto-assigned at session close and can be toggled manually. Test sessions are excluded from ASR calculations.

Is Todo

true when the session requires follow-up action (e.g., callback, relay message).

Contact

The user's name if provided, otherwise "User."

Contact Information

Phone number, email, or other contact details collected during the session.

Summary

AI-generated summary of the conversation.

Lead Value

Estimated revenue if the lead converts. See Lead identification and Lead Value calculation.

Conversion Rate

Percentage of conversations expected to result in a purchase. See Conversion Ratio.

Recognition

Percentage of deals attributed to the AI Employee during business hours. See Recognition.

Session Length

Duration category: Short, Regular, or Long.

Auto assessment

AI-evaluated Agent Success: Yes-A, Yes-B, or No. See Agent success classification and assessment.

Human assessment

Human override of auto assessment. Takes precedence when set. See Agent success classification and assessment.

Conversation Quality

Star-based quality score (CSS). See Scoring metrics: OSS and CSS.

Working Hours

Whether the session occurred during business hours.

Comment

Free-text notes added to the session.

Units

Industry-specific units (e.g., guest count, square footage).

Integration

Integration or connector associated with the session.

External ID

Identifier from an external system (e.g., CRM record).

Actual Result

The outcome that occurred, as determined by AI analysis.

Expected Result

The expected outcome based on user intent.

Session Id

Unique system identifier for the session.

UTMs

UTM tracking parameters from the session source URL.

Filter sessions

Use the search bar and Filters panel to narrow the session list.

The search bar accepts phone numbers and contact names.

Click Filters to open the filter panel. The table below lists all available filters.

Filter

Description

Search

Text search by phone number or contact name.

Is Lead

Filter by lead status.

Is Opportunity

Filter by opportunity status.

Is Test

Filter by test status. Select Live to show only non-test sessions.

Is Todo

Filter by to-do status.

Is Archived

Filter by archived status.

Human Assessment

Filter by human assessment value.

Auto Assessment

Filter by auto assessment value.

Channels

Multi-select: Voice, SMS, Chat.

Session Lengths

Multi-select: Short, Regular, Long.

Integration / Connector

Filter by connected integration.

Date & Time

Preset ranges: Last 24 Hours, Last 7 Days, Last 30 Days.

From Date & Time / To Date & Time

Date pickers for a custom range.

Export a session report

  1. Set your filters (date range, Is Test = Live, etc.).

  2. Click Export Session Report.

  3. Check your email for the CSV file.

The exported CSV includes the following columns: Date/Time, Session Type, Is Lead, Is Opportunity, Is Test, Contact, Contact Information, External ID, Summary, Lead Value, ASR, Conversation Rate, ASR Recognition, Auto Assessment, Human Assessment, Conversation Quality, Working Hours, Comment, Channel, Actual Result, Expected Result, Agent IDN, Units, Integration

Review a session

  1. Click a row to open the session detail panel.

  2. Review the Summary and Transcript.

  3. Check Auto assessment β€” override it with Human assessment if needed.

  4. Add a Comment.

Example: Review a newly launched AI Employee's first week

Suppose you launched an AI Employee for a restaurant called "Sunrise Bistro" one week ago and want to evaluate its performance.

  1. Open the Builder for Sunrise Bistro and click Sessions in the left sidebar.

  2. Click Filters and set Date & Time to Last 7 Days.

  3. Set Is Test to Live to exclude test sessions from the results.

  4. Scan the Auto assessment column. Look for sessions marked No β€” these indicate the AI Employee did not achieve the expected outcome.

  5. Click a No session to open the detail panel. Read the Transcript to understand what happened. If the AI Employee actually handled the call correctly, set Human assessment to Yes-A or Yes-B to override the auto assessment.

  6. Check the Is Lead and Is Opportunity columns to see how many callers expressed purchase intent and how many qualified as opportunities.

  7. Review Conversation Quality stars to identify sessions where the conversation quality was low. Add a Comment to flag sessions that need attention.

  8. Click Export Session Report to generate a CSV of the filtered sessions. Share the report with the Sunrise Bistro owner or use it for internal analysis.

Did this answer your question?