---
title: "Record verdicts after a live coding interview"
description: "Each interviewer records a recommendation and a note on the session's result page, then someone on the team advances or rejects the candidate."
updated: 2026-10-06
canonical: https://kendor.io/docs/live-coding/interviewer-verdicts
---

# Record verdicts after a live coding interview
When a live coding session ends, Kendor keeps the final code and the whiteboard on a result page. Each interviewer picks a **Recommendation** there, and then someone on the team chooses **Advance** or **Reject**.

## Find the result

Ending a session takes you to the candidate's profile. You can also open a result from the live coding interview's **Sessions** list: finished sessions have **View result**. Use **Show sessions** to filter the list to **To decide** or **Advanced**.

The top of the result page shows the candidate's name and email, the challenge, the date, how long the session ran, and where the decision stands: **Awaiting decision**, **Advanced** or **Rejected**.

If a name or email is wrong, for example a typo the candidate made in the lobby, choose **Edit details**. If you change the email, the session moves to that candidate's profile.

## What's kept from the session

The result page has a tab for each kind of evidence.

| Tab | What it shows |
|---|---|
| **Final code** | The project as it was when the session ended. |
| **Terminal** | Terminal sessions, when they were recorded. |
| **AI agents** | Claude Code and Codex conversations, when agents were used and recorded. |
| **Requests** | Requests sent from the editor's Requests panel. |
| **Whiteboard** | A read-only snapshot of the board. |
| **Comments** | A thread for your team. The candidate never sees it. |

To run what the candidate left, choose **Open in editor**. It opens a fresh sandbox from the kept code, so you can start the app or run the tests. Changes you make there aren't saved.

The room's chat isn't kept, and Kendor doesn't record the video call.

## Record your recommendation

In the **Recommendation** card, pick one of **Strong no**, **Lean no**, **Lean yes** or **Strong yes**. It saves right away.

Then add a note to explain your call, and choose **Save note**. The note is optional, and the other interviewers see it. To take your recommendation back, choose **Clear verdict**.

Each person has one recommendation per session, which they can change at any time. Everyone's recommendations and notes are listed under the card, so the panel can compare them.

Owners and editors can record a recommendation. Viewers can read the result page and add comments.

## Make the decision

The **Decision** card shows how far the panel has got:

- **{count} of {required} reviews in. You can still decide.** Not every interviewer on the panel has given a recommendation yet.
- **Every review is in.** Everyone on the panel has given one.
- **The reviewers disagree — talk it through before deciding.** The panel has both yes and no recommendations.

Choose **Advance** or **Reject**, add an optional reason for the decision log, and confirm. You don't have to wait for every review.

Neither choice emails the candidate. If you reject someone, they hear nothing until you send a message yourself.

If you made the wrong call, choose **Undo this decision** to clear it.

> **Tip.** Agree before the session who records the decision, so the panel doesn't have to work it out afterwards.

## Attendance

The **Attendance** card lists everyone who was in the room, with the times they joined and left. For each person it also shows how much they typed, for example "1,240 characters typed in 6 files". It helps you tell how much of the final code came from the candidate and how much from the interviewers.

## Sessions with no result

A session only gets a result page if the candidate joined. If the candidate never joined, a session you started on the spot is removed when you end it, and a scheduled one is marked **No-show** in the **Sessions** list, with no code kept.