AI-Powered Recruitment Admin Automation
- Vacancy prep cut from 45 minutes to one submission
- 5-day delivery, scoping to handover
- 100% consistent CV branding
Project Overview
A two-person UK executive recruitment consultancy was running every vacancy cycle by hand — rewriting notes into advert drafts, manually stripping contact details from CVs before branding them, editing terms templates client by client, and tracking follow-ups from memory. With no orchestration between any of these steps, the Founder was the bottleneck for every document that went out the door. SIDIS built a minimal, n8n-centred pilot that covers the entire admin cycle, from raw notes to ready-to-send output, while leaving every approval with the recruiter.
Challenges to Resolve
Tech Stack Used
From raw notes to ready-to-send documents, in one pass
SIDIS designed and built a minimal, n8n-centred automation pilot covering the full vacancy admin cycle, while every approval decision still sits with the recruiter.
- Vacancy notes to draft pack — raw notes or transcript text go in, and a strict prompt template turns them into a structured vacancy brief and a first-draft job advert. The system checks that role, location and must-have skills are present before producing the draft, so the recruiter never gets a half-finished document.
- CV sanitisation & branding — contact details are detected and redacted automatically, the company header and logo are applied, and a redaction report is generated for every CV. Final export is blocked until the clean check passes.
- Terms & conditions autofill — client name and date populate automatically at the point of trigger, with the output written back to a client-specific SharePoint folder.
- Follow-up reminder automation — the recruiter sets the check-in type and timing once, and dated reminder tasks land in Outlook via the Graph API the moment the vacancy package is processed.
- Action log & audit trail — every run is logged with an ID, timestamp, user and output status, with duplicate CV submissions caught via hash check before they're processed.
AI is scoped narrowly to the steps where turning unstructured notes into structured documents adds real value — every output still goes through human review before it reaches a client.
Why It Matters