Who can use this: Any team member
Available on: Enable Essentials, Enable
Ideal for: Sales manager, enablement, QA reviewers
AI Call Reviewer is a Gong AI agent that answers scorecard questions from the call transcript. An admin turns it on for your company and marks which scorecard questions use AI. AI Call Reviewer then either suggests answers while a reviewer scores a call or scores matching calls automatically. Its results appear on the call page alongside manually submitted scorecards.
How AI Call Reviewer works
AI Call Reviewer contributes to scoring in two ways for scorecard questions configured with Get answer from AI.
AI-suggested answers (manual review)
When a reviewer clicks Scoring (the star icon) on the right side of a call page, AI Call Reviewer shows a suggested answer for each AI-enabled question. The reviewer can accept the suggestion, change it, or override it before submitting. The reviewer is considered the scorer of the submitted result.
Automatic review
When a scorecard is set to automatic review, AI Call Reviewer evaluates every matching call after it is recorded and processed, and generates a completed scorecard result, so no one needs to open the call. Automatic results are labeled Scored by Gong AI on the call page. Learn more about reviewing automatically scored calls.
Note:
Automatic review must be enabled at the company level by a tech admin in Admin center before any automatic scorecards score calls. Learn more about setting up AI Call Reviewer.
What AI Call Reviewer can and can't do
AI Call Reviewer has these limits:
It cannot score voicemail calls or calls with fewer than 100 words in the transcript.
It only answers questions that have been configured with AI assistance. Questions without an AI source are left blank for the reviewer to complete manually.
Automatic scores do not count as deliberate manager coaching in Coaching metrics. Those views are designed to reflect intentional coaching actions such as listening, commenting, and manually scoring calls.
Where results appear
AI Call Reviewer results appear in the same places as manually submitted scorecards:
On the call page, in the Scoring panel (the star icon). Automatic results are labeled Scored by Gong AI.
In Insights > Team > Scorecards, where average scores and trends are visible across the team.
In CSV exports from scorecard management or Team Insights.
Visibility rules apply to AI-generated results the same way they apply to manual results. If the scorecard's visibility is set to Scorer, Gong AI is treated as the scorer. The reviewed person or their manager may then not see the result on the call page.
What you need
To use AI Call Reviewer:
For automatic review, a tech admin turns on Allow automatic review for the company in Admin center > Settings > General settings > AI Call Reviewer. Learn more about managing general company settings.
A scorecard with AI-enabled questions exists. For automatic review, every question uses Get answer from AI.
Once automatic review is enabled, scorecards configured for automatic scoring start evaluating new calls after they are recorded and processed. You can also apply an automatic scorecard to calls from the past 90 days using Apply to past calls.
Frequently asked questions
Does an automatically scored call count as feedback given in my Coaching metrics?
No. Automatic scorecard results do not count as deliberate manager coaching. Coaching metrics and Coaching inbox are designed to reflect intentional actions: Listening, attending, manually scoring, commenting, and marking offline feedback. Automatic scoring tracks whether a call was evaluated against your criteria, separately from whether a manager personally coached it.
Can I see AI-suggested answers before submitting a scorecard?
Yes. When you open the scoring panel for a call, questions with AI assistance show a suggestion. Click AI answer to read the AI's rationale and the transcript evidence it used. You can keep the suggestion or enter a different answer before submitting. The final submitted result is yours, not the AI's.
What's the difference between AI-suggested answers and automatic review?
AI-suggested answers assist a human reviewer. AI Call Reviewer proposes an answer, but the reviewer must open the call, review the suggestion, and submit the scorecard. The reviewer is the scorer. Automatic review requires no human action at all: AI Call Reviewer generates and submits the result for every matching call. Automatic results are labeled separately in all views so you can always tell which results are AI-generated.
Why do I see different AI scores for the same call?
AI sometimes produces different outputs for identical inputs. If you review the same call again or apply an automatic scorecard to a previously processed call, the suggested answers may differ slightly. The underlying call transcript hasn't changed. The variation is a normal characteristic of how large language models work.