To get a report in Interviewer Mode, click 📊 Evaluate Answer after each answer you want scored, then click 📊 Summary. CoPilot Interview writes a candidate evaluation with the questions asked, an overall score out of 10, a recommendation of Hire, No hire or Maybe, scores with evidence in five fixed areas, gaps with practice drills, strengths, likely next-round questions and important topics you didn't cover. It's a draft for your judgment, not a decision.
CoPilot Interview is an AI interview assistant for Windows and macOS, and Interviewer Mode is its mode for people running interviews. For the setup screen, see how to set up Interviewer Mode.
How to generate the report
- During the session, click 📊 Evaluate Answer after each candidate answer you want in the report. Questions without an evaluated answer can be listed, but they aren't scored.
- Click 📊 Summary in the session bar. If nothing has been evaluated yet, the app shows “No candidate answers recorded yet” and asks you to use 📊 Evaluate Answer first.
- The report appears as a card titled 📊 Candidate Evaluation.
- Click End. The report is saved under “AI candidate evaluation” in the session's Markdown file in Documents › CoPilot Interview › Sessions.
What's in the report
| Section | What it contains |
|---|---|
| Header | 📊 Candidate evaluation, with the interview type, the role and company if they were stated, the number of questions and the minutes |
| 📋 Questions asked | Up to 30, in order, each with a neutral title, category, difficulty and what it was testing |
| 🎯 Scorecard | Recommendation: Hire, No hire or Maybe · Overall score out of 10 · a short verdict with the main reason |
| Five-area table | Dimension, Score and Evidence for each of the five areas below |
| 🔮 Likely next-round questions | 5 to 8 questions that probe the gaps and any unverified claims |
| 🛠 Gaps to work on | 3 to 6 gaps, most important first, each with a concrete drill the candidate could use |
| 💪 Strengths | The candidate's top 3 to 5 |
| 🧭 Not covered | 2 to 6 important areas for this role that weren't asked about |
The five scored areas
- Content & accuracy
- Structure (STAR/clarity)
- Communication
- Problem-solving
- Role fit
Each area gets a whole number from 1 to 10, with evidence that quotes or paraphrases what the candidate said. An area shows “—” when it can't be judged from what was captured. The recommendation comes from the evidence only, and the AI is told to choose Maybe when the evidence is thin. Section headings, category names, area names and the recommendation stay in English; the rest follows your 🌐 Lang setting.
Limits to know
- Fixed areas. The report scores the same five areas every time, in whole numbers. It doesn't create per-topic scores or custom rubric dimensions; add your competencies to the topics box to steer the questions and evidence.
- Only what was captured. Scores rest on the transcript. If the candidate's audio was weak, the evidence is too.
- AI-written. Read it against your own notes before you rely on it, and remove anything the evidence doesn't support.
- ⚡ Dual AI Mode. With it on, both models write a report, side by side.
- Cut-off reports. If you see “did not return a complete report — its reply was cut off twice. Press 📊 Summary again, or pick another model for this lane”, do exactly that.
- Anonymous stats. A line under the report lets you turn off anonymous stats. They are on by default except in the EEA, the UK and Switzerland, and never include answers, transcripts, audio, résumés, names, email addresses or license keys.
Use the report fairly
- Ask every candidate for a role the same core questions, using 📝 Specific Questions, so their reports compare like for like. Comparing candidates fairly with AI scorecards goes further.
- Keep your own written scorecard and treat the report as a second opinion; how to write an interview scorecard shows what to include.
- Use 🧭 Not covered to plan the next round, and 🔮 Likely next-round questions as a starting list.
- If your policy allows candidate feedback, the drills under 🛠 Gaps to work on are written so they can be shared.
Frequently asked questions
Can I export an Interviewer Mode report?
Every session is saved automatically as a Markdown file in Documents › CoPilot Interview › Sessions, and the report is in it under “AI candidate evaluation”. Open the file to copy or share the report.
Why does one area show a dash instead of a score?
The AI couldn't judge that area from what the candidate said. Ask a question that targets it, evaluate the answer, and run 📊 Summary again.
Can I change the five scoring areas?
No. The areas are fixed: Content & accuracy, Structure (STAR/clarity), Communication, Problem-solving and Role fit. You can steer what is asked through the job description, topics and your own questions.
Does the candidate see the report?
No. It's created and saved only on your computer. Share it only if your process allows, after checking it.
Run your next interview in Interviewer Mode
Suggested questions from the job description, a score for each answer you evaluate, and a candidate evaluation with a Hire, No hire or Maybe recommendation. Same desktop app for Windows and macOS, free plan included.
Get the desktop app