Case Study · Recruitment Agency

How a Mid-Sized Recruitment Agency Cut CV Screening Time by 80% with AI Agents

A 12-person recruitment agency was spending over 30 hours a week manually screening CVs, chasing candidates, and updating their ATS. We built them a fully automated hiring pipeline on n8n — deployed in under a week.


80%
Reduction in time spent screening and shortlisting candidates
3x
More placements made per month with the same headcount
£0
Upfront cost — paid only after the solution was live and working

Recruiters were spending most of their day on admin, not placements

The agency was receiving 100–200 CVs per open role. Each one had to be manually opened, read, scored, and either rejected or moved forward. This process alone was consuming 3–4 hours per recruiter per day.

Candidate follow-ups were being done manually via email. Interview confirmations, rejection notices, and status updates all required a human to write and send them. Candidates were regularly left waiting days for a response, damaging the agency's reputation.

Their ATS was being updated inconsistently — data was missing, stages were wrong, and managers had no reliable view of the pipeline. The team was busy but the output wasn't matching the effort.

A fully automated candidate pipeline built on n8n

We mapped the agency's existing workflow and identified every manual touchpoint. We then built a set of n8n AI agents to handle each one — with full error handling, idempotent operations, and real-time ATS sync.

  • AI CV Screener: Every inbound CV is parsed, scored against the job spec, and ranked automatically. Recruiters only see the top-matched candidates.
  • Automated Candidate Comms: Acknowledgement emails, interview invites, and rejection notices are sent automatically based on pipeline stage — personalised using the candidate's name and role.
  • ATS Auto-Update: Every action triggers a live update to the ATS. No manual data entry. Pipeline stages, notes, and timestamps are always accurate.
  • Recruiter Daily Digest: Each morning, recruiters receive a prioritised summary of who to call, who's awaiting a decision, and which roles are most urgent.
  • Interview Scheduling Agent: Candidates are sent a self-booking link. Confirmations and reminders are handled automatically, including rescheduling logic.

Before & After

AreaBeforeAfter
CV screening time per role8–10 hours manual reviewUnder 20 minutes, AI-ranked shortlist
Candidate response time2–4 days averageUnder 30 minutes, automated
ATS data accuracy~60%, updated inconsistently100%, updated in real time
Monthly placements8–10 per month24–28 per month
Admin hours per week30+ hours across the teamUnder 6 hours
Recruiter focusMostly admin and data entryClient relationships and closing
CV screening time per role
Before8–10 hours manual review
AfterUnder 20 minutes, AI-ranked shortlist
Candidate response time
Before2–4 days average
AfterUnder 30 minutes, automated
ATS data accuracy
Before~60%, updated inconsistently
After100%, updated in real time
Monthly placements
Before8–10 per month
After24–28 per month
Admin hours per week
Before30+ hours across the team
AfterUnder 6 hours
Recruiter focus
BeforeMostly admin and data entry
AfterClient relationships and closing

The same team. Three times the output.

Within two weeks of going live, the agency had cleared a backlog of 6 open roles that had been stalled for months. Recruiters reported spending their time on calls and client meetings instead of spreadsheets and email drafts.

The agency now runs a leaner, faster hiring operation without adding a single headcount. The AI agents work around the clock — screening overnight applications, sending morning follow-ups, and keeping the ATS clean without anyone lifting a finger.

This was built, tested, and deployed in 5 days. The agency paid nothing upfront.

Book your free AI audit →