Desktop Sessions

Work from a single record of each call

Sessions are the main record inside Close AI. Every live interaction in the desktop app creates a session, whether you use Listen mode for live coaching from call audio or Ask mode to question the AI during a call.

When the call ends, the session stays available for review. You can reopen it to check the transcript, inspect AI messages, read the summary, and run analysis for coaching and script coverage.

Find and scan sessions quickly

The sessions page shows a searchable, paginated list of desktop call records. Each page displays up to 25 sessions, so reps can review recent calls while managers can scan activity across the team.

What each session card shows

Each session card surfaces the details you need to decide whether to open it:

• Title — the session name, with a fallback based on the session ID if no custom title exists
• Session type — Listen or Ask
• Status — active, ended, completed, or another recorded state
• Started — the session start time
• Duration — elapsed call time
• Short ID — the first part of the session ID for quick reference

Session status indicators

Desktop status banner

At the top of the list, the desktop status banner shows whether the workspace has desktop access and how much monthly usage remains.

Open a session to review the full call

Select a session to open its detail view. The header gives you the basic facts first so you can orient quickly before digging into the call.

Header details

The top of the session detail page includes four metadata cards:

You can also rename the session inline using the pencil edit control. If the current title is too generic, update it to match the customer, deal, or objection discussed on the call.

Delete a session

Delete removes the session record from the list. Use this when a session was created by mistake or should not stay in the workspace.

Use the summary card for a fast recap

If AI summary data exists, the session shows a summary card near the top of the detail view. This is the fastest way to understand what happened without reading the entire transcript first.

The summary card can include:

• TL;DR — a short written recap of the call
• Key points — the main topics or moments pulled into bullet points
• Action items — follow-up work captured from the conversation
• Token usage — how much model usage the summary consumed

For managers, this is often enough to spot whether a call needs deeper coaching. For reps, it gives a clean handoff into follow-up work.

Run analysis for coaching and review

The analysis section turns a completed session into a coaching record. Use Analyze to generate analysis for the first time, or Re-analyze to refresh the results after transcript data changes.

Talk metrics

These metrics show how the rep handled the conversation at a speaking-pattern level.

CloseScore breakdown

CloseScore gives the session an overall score from 0 to 100 and breaks that score into categories. The score is color-coded so you can judge performance at a glance.

Below the overall score, the analysis shows category-level breakdowns with sub-items and notes. That helps managers see whether the score dropped because of discovery, handling objections, pacing, or script execution rather than treating the number as a black box.

Strengths and improvements

Coaching results are split into two practical sections:

• Strengths — behaviors the rep should keep repeating
• Improvements — weaknesses and risks that need correction

This layout makes post-call review faster because the rep can see both what worked and what to change in one pass.

Script coverage

If the session used a script, analysis also shows how fully the conversation followed it. The checklist marks each section based on coverage and shows the time spent in that section.

You can use script coverage to answer questions like:

• Did the rep reach the pricing section
• Did the call skip a key discovery step
• Where did the conversation spend the most time

Prospect personality

The personality block helps the rep adapt follow-up and future calls. It shows:

• Personality type
• Levers — what tends to motivate this prospect
• Avoid — what tends to irritate or disengage them

Review the transcript and AI message history

Below the summary and analysis, session content is organized into tabs so you can switch between the call itself and the AI's internal conversation.

Transcript tab

The transcript tab shows speaker-attributed rows with timestamps and language tags. The layout uses chat-style bubbles, which makes it easier to follow turn-taking during longer calls.

This view is the best place to verify exactly what was said before giving coaching feedback or writing a follow-up.

AI Messages tab

The AI Messages tab shows the message log behind the session. Entries are grouped by role:

• System
• User
• Assistant

Each message includes its role, token count, and time. Use this tab when you want to understand how the AI responded during the session or troubleshoot why guidance looked a certain way.

Use the desktop dashboard for team trends

Individual sessions help with call-by-call review. The desktop dashboard helps managers step back and see patterns across the team.

KPI cards

The dashboard header summarizes performance for the selected time range.

Time ranges and analysis tabs

Use the time selector to switch between 7d, 30d, and 90d views. Each range updates the charts and tables across the dashboard.

The dashboard includes eight analysis tabs:

AI narrative insights

Each dashboard tab can generate a written narrative based on the current data. This adds a short AI summary and recommendation layer on top of the charts.

For managers, the narrative is useful when preparing coaching reviews or team updates. For reps, it helps turn dashboard trends into a concrete next action.

Related pages