Interview Analysis

Transform raw interview notes into a structured, paste-ready summary. Scores each question against the rubric, synthesizes a recommendation, and generates interviewer self-coaching for improvement.

When to use

• "Clean up interview notes", "analyze interview for [candidate]", "interview analysis"
• "Score this interview", "write up my interview notes"
• After conducting a behavioral or values-based interview

Inputs

The user must provide:

1. Raw interview notes — the interviewer's real-time notes. Can be:
   • Pasted directly into the conversation
   • A file path to the notes document
   • Already in a candidate file that needs cleanup
2. Interview template/rubric — the question bank and scoring criteria. Can be:
   • A file path to the template
   • Pasted directly
   • Referenced by name if the user has a known templates directory
3. Target level (optional) — helps calibrate rubric expectations. Infer from template name if not stated.
4. Interviewer name (optional) — for attribution in the summary header. If not provided, omit or ask.

If any required input is missing, ask the user to provide it before proceeding.

Output Structure

Produce a clean summary followed by the preserved raw notes:

# Behavioral Interview Summary: [Candidate Name] ([Level] Candidate)

**Candidate:** [Name] **Target Level:** [Level] **Interview Type:** Behavioral **Interviewer:** [Name]

**Recommendation: [Hire / No Hire / Lean Hire / Lean No Hire]**

**Overall Impression:** [2-4 sentence synthesis of the interviewer's assessment. Direct and honest.]

**Key Concerns:** (if applicable, bullet list of specific patterns observed)
**Key Strengths:** (if applicable, bullet list of standout positives)

---

## Question N: [Topic Name]

**Q:** _[Full question text from the template, italicized]_

**Values/Competencies:** [Dimensions from the template for this question]

**Summary:** [Narrative paragraph synthesizing the raw notes into coherent prose. Include what was asked, what the candidate said, how follow-ups went, and how the candidate responded to probing. Tell the story of the conversation, not just summarize the answer.]

**Lesson Learned:** [Include ONLY when the question is about failure, feedback, or growth. Extract what the candidate claims to have learned. Omit for other question types.]

**Org Fit Signal:** [Optional. Include when the answer reveals signal about what kind of role, team, or environment the candidate would thrive in. Omit if no clear signal.]

**Rubric Assessment:**

- [✅/⚠️/❌] [Rubric criteria from template]; [brief evidence or gap noted]
- [✅/⚠️/❌] [Next rubric criteria]; [brief evidence or gap noted]
...

---

[Repeat for each question asked]

---

## Candidate's Questions

- [List questions the candidate asked]
- [If the questions reveal signal, add a brief note]

---

## Interviewer Notes

[Synthesis of behavioral observations, communication style, red flags, and patterns. Written in first person from the interviewer's perspective. Professional but direct.]

---

## Interviewer Self-Coaching

For each question asked, note what follow-ups could have been asked to get stronger signal.

### Question N: [Topic]
- **Missing signal on [area].** A good follow-up would have been: _"[suggested question]"_ This would test [what it would reveal].

### General Notes
- [Overall interviewing observations: pacing, depth vs. breadth, time management]

---
---

# Raw Interview Notes

[Original raw notes preserved exactly as written]

Rubric Assessment Rules

For each question, pull rubric criteria from the template (both Positives and Negatives columns if available). Assess each positive criterion:

• ✅ Strong positive signal: candidate provided specific, concrete evidence
• ⚠️ Mixed or partial signal: candidate touched on this but lacked specifics or depth
• ❌ Negative signal or absent: candidate did not address this, or actively demonstrated the corresponding negative

Format: - [emoji] [Criterion from template]; [your specific evidence or observation]

Use semicolons (not colons or dashes) to separate criterion from evidence. When a candidate's answer matches a negative from the rubric, call that out explicitly.

Determining Recommendation

Read the raw notes for explicit hire/no-hire signals from the interviewer. If stated, use that. If not explicit, derive from rubric assessments:

• Majority ✅ with no ❌ on critical criteria = Hire or Lean Hire
• Mixed ✅ and ⚠️ = Lean Hire or Lean No Hire depending on severity
• Any ❌ on critical criteria (depth of expertise, decision-making, learning from failure) = Lean No Hire or No Hire
• Multiple ❌ or clear behavioral red flags = No Hire

How to Identify Questions Asked

Match questions in raw notes against the template's question bank by content/topic. Note which template question number each corresponds to. If a question doesn't match the template, include it as supplemental.

Follow-Up Conversation Flow

The Summary for each question should document the full conversation arc:

• What the candidate said initially
• What follow-up questions were asked
• How the candidate responded to probing (deeper, redirect, repeat, struggle?)
• Whether hints or reframes were needed
• Editorial observations from raw notes woven in naturally

Handling the Raw Notes

• Preserve raw notes exactly as written below the double separator (---\n---)
• Do not fix typos or grammar in the raw notes section
• The summary section is the polished version; raw notes are the source of truth

Rules

• Self-contained. Do not reference other candidate files as examples or calibration.
• Always generate the Self-Coaching section. For each question, suggest 1-2 follow-ups that could have yielded stronger signal. Reference specific rubric criteria that were hard to assess. Be constructive and specific, not generic.
• Preserve existing content. If writing to a file, raw notes stay intact below the separator.
• Be honest in assessments. Don't grade on a curve. If signal is weak, say so.
• Output flexibility. If the user doesn't specify a file path, present the analysis in the conversation. If they specify a path, write it there.