A technical headhunting firm was spending 40 minutes reformatting every CV before it could reach a client. We built an AI pipeline that parses, anonymizes and rebuilds each CV as a branded profile in three minutes flat.
Typical results
30
Weekly submissions per consultant
0
Formatting fixes needed
100%
On-brand submissions
Figures reflect typical results for this kind of automation.
The challenge
Every submission required a consultant to copy CV content into the firm's template, strip out contact details, rewrite the summary and fix the formatting — roughly 40 minutes per candidate. On a busy search with ten candidates, that was a full afternoon lost, and rushed profiles went out with typos, stale data and inconsistent branding.
Before and after
Prep time per profile
−93%Weekly submissions per consultant
3.8×Formatting fixes per month
−100%The automation
We built a Make scenario around OpenAI that triggers the moment a CV is uploaded or forwarded by email. AI extracts experience, skills and education, strips every identifying detail, and writes a sharp two-paragraph summary in the firm's voice. The data flows into the branded profile template, the consultant gets a preview link for a one-click approval, and the finished PDF files itself to the ATS, ready to send.
How it works
- 01
CV arrives
A consultant uploads a CV or forwards it by email; the Make scenario picks it up instantly.
- 02
AI parses & anonymizes
OpenAI extracts experience, skills and education, and strips every identifying detail from the text.
- 03
Template filled
The structured data and an AI-written summary flow into the firm's branded profile template.
- 04
One-click approval
The consultant reviews a preview link and approves — or edits — in seconds.
- 05
Filed & sent
The finished PDF is saved to the ATS and delivered to the client, fully on brand.
Starts from
- CV uploads
- Email attachments
Delivers
- Branded profile PDF
- ATS record
- Client submission
The impact
Profile preparation dropped from 40 minutes to 3, and weekly submissions per consultant jumped from 8 to 30. Formatting fixes — previously around 15 a month — disappeared entirely, and every profile now leaves the firm in identical, polished branding. Consultants spend the reclaimed hours talking to candidates instead of wrestling documents.
Questions about this automation
How does the anonymization step work?
After OpenAI extracts experience, skills and education into structured fields, the automation removes names, phone numbers, email addresses, links and other identifying details from the text. Only the cleaned fields reach the branded template, and the consultant reviews the preview before anything is sent to a client.
Can the profile template match our existing branding?
Yes. The template is built from the firm's current profile design, with its logo, fonts, colors and section order. The AI-written summary follows a short style guide for tone and length, and there can be several template versions, for example one per practice area or seniority level.
What if the AI misreads a CV or gets a detail wrong?
Nothing goes out without the consultant's one-click approval. The preview link shows the finished profile next to the original CV, and any field can be edited before approving. Unusual formats such as scanned images or complex tables are flagged for a closer look instead of being filled in silently.