Turn recordings into scorecards automatically

Founder or hiring manager at a 10-80 person startup running 5-30 interviews per month without a dedicated recruiter · Assemble in ~10 min; first scorecard from a recording in <5 min once installed.

HiringEdit flowon X● golden 91

Not legal or HR-compliance advice. Employment and hiring rules vary by jurisdiction — have qualified counsel review policies, offers, and screening before use.

step 1ready

Define the role and scoring rubric

interview-kit1 alts

step 2connector

Obtain the recording or transcript

⚡ bring your own tool

step 3ready

Analyze the transcript for behavioral evidence

meeting-analyzer

step 4ready

Score the candidate and generate the structured notes

ent-interview-debrief

step 5ready

Produce the shareable debrief document

meeting-notes6 alts

step 6partial

File the scorecard and update the hiring pipeline view

hiring-pipeline-report3 alts

step 7ready

Synthesize patterns across a batch of candidates

ent-synthesis-coach

Did this recipe work for you?

What you get

▸Completed per-candidate scorecard with competency ratings and evidence quotes
▸Structured interview notes with facts-vs-interpretation separation and flagged red flags
▸Shareable debrief document ready for the hiring panel or ATS upload
▸Cross-candidate synthesis and updated role bar after each interview loop

tools: Transcript source: Fathom free tier / Otter.ai ~$17/mo / Fireflies ~$10/mo; ATS optional

tokens: ~$0.30-0.80 per interview (60-90 min transcript)

≈ $10-25/mo at 20 interviews/month

What stays yours

· Review the hire/no-hire recommendation — the scorecard informs, the decision is yours; employment decisions carry legal weight and should not be delegated to AI output alone.
· Calibrate the competency rubric after the first 5-10 scorecards to ensure dimensions reflect what actually predicts success in the role.
· Conduct the debrief conversation with the panel — the document structures it, but alignment happens in the room.

✓ Last verified 2026-06-11claude-sonnet-4-6 (subagent, test-mode fixtures)

✓ step 1 · Define the role and scoring rubric

✓ step 2 · Obtain the recording or transcript

✓ step 3 · Analyze the transcript for behavioral evidence

✓ step 4 · Score the candidate and generate the structured notes

✓ step 5 · Produce the shareable debrief document

✓ step 6 · File the scorecard and update the hiring pipeline view

✓ step 7 · Synthesize patterns across a batch of candidates

view sample outputs from this run

step 1 · Define the role and scoring rubric

# Role Scorecard Template
## Head of Growth — AI-Native B2B SaaS (Series A)

**Role:** Head of Growth  
**Team:** Reports to CEO; 0→1 growth function  
**Interview loop:** Screening call → Case study → Structured loop (this template) → Reference check  
**Scorecard owner:** Hiring manager (CEO)  
**Fixture candidate:** Maya Chen, 7 years experience, ex-Notion, ex-Amplitude  

---

## Scoring Scale

| Score | Label | Meaning |
|-------|-------|---------|
| 4 | Exceptional | Exceeds bar; would be one of our best at this |
| 3 | Strong | Meets bar with evidence; minor gaps acceptable |
| 2 | Deve…

step 3 · Analyze the transcript for behavioral evidence

# Behavioral Evidence Map
## Candidate: Maya Chen — Head of Growth Interview
## Mapped against: competency-rubric.md

---

## Summary

Transcript parsed for behavioral anchors across all 7 competency dimensions. Evidence is quoted verbatim or paraphrased from the transcript with location references. Interpretation is explicitly labeled.

---

## Dimension 1: PLG Execution

### Evidence Found

**E1.1 — Amplitude activation funnel redesign (strong positive)**
> "Trial signups were healthy but only 19% of users reached what we called the 'first chart moment'... we built what we called a sample-da…

step 4 · Score the candidate and generate the structured notes

# Completed Scorecard
## Candidate: Maya Chen — Head of Growth
## Interviewer: Stepan (CEO)
## Interview date: 2026-06-09
## Scorecard completed: 2026-06-09

---

## Dimension Scores

| # | Dimension | Weight | Raw Score | Weighted |
|---|-----------|--------|-----------|---------|
| 1 | PLG Execution | 20% | 4 | 0.80 |
| 2 | Data & Experimentation Rigor | 20% | 4 | 0.80 |
| 3 | Founder-Mode Hustle | 15% | 4 | 0.60 |
| 4 | GTM & Demand Generation | 15% | 3 | 0.45 |
| 5 | Cross-functional Influence | 10% | 3 | 0.30 |
| 6 | Customer Insight | 10% | 4 | 0.40 |
| 7 | Communication & Clarity | 10% …

The problem

When I finish interviewing a candidate, I want a structured scorecard and summary auto-generated from the recording, so that I can make consistent, bias-reduced hiring decisions without spending 40 minutes per interview writing notes.

Cybernetic layers

Memorybyo

Candidate history, past scorecards, and role-bar evolution live in your ATS or a shared doc — not yet a library artifact. Use Notion, Google Sheets, or a local context file to accumulate scorecards across candidates.

Backgroundbyo

No automated trigger yet — each scorecard run is human-initiated after the meeting ends. A Zapier/Make trigger on new Fathom/Otter transcript could kick off the pipeline automatically, but that connector is byo.

Harnessready

Claude Code is the runtime. Drop each skill into ~/.claude/skills/ and invoke by name. The full batch runs in a single session per interview.

Click any step to see what fills it — and the ranked alternatives you can swap in.