How to convert a screening call transcript into a client-ready CV
Half your candidate intake doesn't come as a clean CV anymore. It comes as:
- A 30-minute Otter or Fireflies transcript from a screening call
- A Gong recording of a discovery conversation
- A LinkedIn message with experience pasted into the body
- A Fathom-recorded networking call from RecFest
The recruitment industry's tools haven't caught up to this reality. Every CV formatter on the market — except FormaCV — assumes you have a clean PDF or DOCX as your input. Which means recruiters spend 30-45 minutes after every call writing the candidate up by hand into a CV before they can send anything to a client.
This post covers the workflow for getting from "I just hung up the call" to "the client has a polished CV in their inbox" in under 5 minutes.
When this matters
This workflow is essential for:
- Executive search firms — partner-led discovery calls are the standard intake; CVs come later, if ever
- IT and tech staffing — technical screening calls produce richer signal than written CVs
- Boutique agencies — informal coffee chats and referrals where no CV exists yet
- RPO and high-volume staffing — phone screens that need to be turned around same-day
If your pipeline is 100% job-board CVs, this workflow doesn't apply. If any meaningful portion of your candidate intake comes through calls, this is your missing tool.
The 5-step workflow
Step 1: Record the call (with consent)
Use whichever tool fits your stack:
- Otter.ai — best transcription accuracy for English; recorder + transcription in one
- Fireflies.ai — good integration with Salesforce / HubSpot
- Gong — sales-focused but works for recruiter calls
- Fathom — free for individuals; good Zoom integration
- Read.ai — combined transcription + meeting analytics
- Microsoft Teams transcription — built-in if you're on Teams
- Zoom transcription — built-in if you're on Zoom
Consent: Most jurisdictions require informing the candidate the call is being recorded. UK and EU GDPR rules require explicit consent; US varies by state (some are one-party-consent). Always inform at the start of the call.
Step 2: Get the transcript text
Most transcription tools auto-generate the transcript within 5-10 minutes of the call ending. Export options:
- Otter: Export → Plain text
- Fireflies: Settings → Export → TXT or DOCX
- Gong: Library → Call → Transcript → Copy
- Fathom: After-call email contains the full transcript
- Teams: Recording → Transcript → Download
What you want is text. Audio files don't help — every CV formatter (including FormaCV) processes text, not audio.
Step 3: Process the transcript
This is the step where every other CV formatter on the market falls down. They expect a CV file as input.
With FormaCV, paste the transcript directly into the upload box. The AI:
- Identifies the speakers (you and the candidate)
- Extracts the candidate's facts: name, current role, employer, experience timeline, skills, education, location, salary expectation, notice period
- Filters out interview filler (the introductions, the "uhms", the recruiter questions)
- Structures the content into your branded CV template
This typically takes 60-90 seconds.
With other formatters, you'd need to:
- Manually re-write the transcript into a CV format yourself (25-40 minutes)
- Then run that hand-written CV through the formatter (60 seconds)
The 25-40 minutes of manual write-up is the bottleneck. That's what FormaCV's transcript handling eliminates.
Step 4: Review and edit
The AI extracts what was discussed. It doesn't make things up — if the candidate didn't mention their education, the education section will be sparse. If they didn't list specific tools, the skills section will be light.
Review what came out. Common edits:
- Add information you remember from the call that wasn't in the transcript (the candidate's tone, energy, specific examples — keep these as your private notes, not in the client-facing CV)
- Reorder bullet points by client relevance
- Polish phrasing where the candidate spoke informally
Total review time: 3-5 minutes for a 30-minute call.
Step 5: Send to client
Once reviewed, the CV exports as DOCX or PDF — same as any other FormaCV output. Send via your normal client submission channel: ATS push-back, email attachment, branded landing page (if you use HireAra alongside FormaCV for that — they coexist fine).
Total time from call ending to client receiving CV: typically under 5 minutes.
What gets extracted from a typical screening call
A 30-minute recruiter screening call usually contains enough signal for FormaCV to extract:
| Field | Source in transcript |
|---|---|
| Name | Speaker introduction, "I'm..." |
| Current role and employer | "I'm currently a senior engineer at X", "I work at Y as their head of..." |
| Experience timeline | "Before that I was at...", "I joined Z in 2021..." |
| Skills and tools | "We use Python and AWS", "I've spent the last two years on..." |
| Education | Often comes from a direct question; sometimes self-introduced |
| Salary expectation | Salary discussion (kept in your private notes, anonymized in client submission output) |
| Notice period | "I'd need to give 3 months", "I could start after..." |
| Location and remote preference | "I'm based in...", "I'm open to..." |
| Soft signals | Communication style, energy, structured response patterns — captured as recruiter-only notes, not in the candidate-facing CV |
A typical 30-minute call yields enough text for a 1.5-2 page candidate profile.
What doesn't transfer well from transcripts
- Quantified achievements — candidates rarely volunteer specific numbers in casual conversation. If a metric matters for the client, ask explicitly during the call.
- Project details — candidates skim over projects in calls; if you need depth, ask follow-up questions
- Visual portfolio elements — design, code, writing samples — these need separate links in the CV, not derived from the transcript
These gaps aren't a FormaCV limitation — they're a limitation of the call format itself. Even a manual write-up from a 30-minute call has the same gaps.
Real-world results from executive search firms
Boutique retained search firms running 12-15 partner-led discovery calls per month see:
- Write-up time: 35-50 min/call → 5-7 min/call (85% reduction)
- Time-to-client-submission: 4-7 days post-call → 24h post-call
- Partner/associate hours reclaimed: ~30 hours per month per firm
The reclaimed time goes to two places: more discovery calls (more candidate volume), or more client conversations (more assignment volume).
Frequently asked questions
Does FormaCV transcribe audio or video?
No — FormaCV processes the text transcript output of your existing transcription tool. We don't compete with Otter, Fireflies, or Gong; we work with their output.
Which transcription tool produces the best results for FormaCV?
All major tools produce text quality FormaCV can process. Otter and Fireflies tend to have the highest English transcription accuracy. Internal benchmarks: Otter ~95% word accuracy on clean English audio; Gong and Fathom similar.
How accurate is FormaCV's extraction from a transcript?
We benchmark internally at ~92% accuracy on standard recruiter screening calls. You always review and edit before sending. The 8% inaccuracy is usually in subtle phrasing (e.g. swapping "led" for "managed") rather than factual errors.
Can I use FormaCV with a Microsoft Teams meeting transcription?
Yes — Teams produces standard transcript text. Copy from the Teams transcript view and paste into FormaCV.
What about consent and GDPR?
You're responsible for compliance with consent rules in your jurisdiction. FormaCV processes transcript text under our standard GDPR-compliant data policy: encrypted in transit and at rest, EU/US data residency, transcript text deleted within 90 days of processing (configurable on Agency+ plans), zero AI training on customer data.
Does the transcript need to be in any specific format?
Plain text works best. Most tools export plain text by default. Speaker labels (e.g. "Recruiter:" / "Candidate:") help FormaCV identify whose information to extract, but aren't required — the AI infers from context.
Can I do this with HireAra, Allsorter, or Daxtra?
To our knowledge, none of those tools currently support transcript-to-CV conversion. They require a CV file as the starting point. If transcript handling is essential for your workflow, FormaCV is the only option in the recruitment-formatting market today.
What if my candidate's transcript is in French / Spanish / German?
FormaCV currently supports English, Spanish, French, German, Dutch, and Italian. Other languages are on the roadmap. For other languages, transcribe and translate first, then process the English transcript.
Try FormaCV with a sample transcript
We've put a sample call transcript on our call-transcript-to-CV feature page. Paste your own transcript or use the sample — see the structured CV output without signing up.
If it fits your workflow, start a 7-day free trial. No credit card.