◉ HASSELT · BEBook →
← CASE STUDIES·HR · BE

Fenix Staffing screens 4× more candidates with the same team.

An autonomous agent that reads every CV, scores it against the role rubric, ranks the applicant pool, and books interviews with top candidates, before a recruiter opens their inbox.

-82%
SCREENING TIME PER ROLE
CANDIDATES REVIEWED PER RECRUITER
11
DAYS TO PRODUCTION
/ THE SITUATION

Recruiters were reading hundreds of CVs to shortlist twelve.

High-volume roles attracted 150–300 applications. Recruiters spent the first two days of every hire manually reading CVs, most of which were obviously mismatched within 30 seconds. By the time the shortlist was ready, the best candidates had already accepted offers elsewhere. Speed-to-shortlist was the constraint, not recruiter skill.

/ WHAT WAS BUILT

An autonomous screening agent that delivers a ranked, annotated shortlist, automatically.

Recruiters define the scoring criteria once per role. The agent reads every CV, scores and annotates each candidate, ranks the pool, and books interviews with the top tier, all before the recruiter's first coffee.

01

Role criteria structured

For each open role, recruiters define scoring dimensions, required skills, seniority, location, industry background, in a simple template. The agent uses this as its evaluation rubric.

02

CV ingestion pipeline built

Incoming applications parsed from email attachments and the ATS. CV content extracted and normalised, regardless of format, layout, or language, before scoring.

03

Autonomous scoring and ranking

Agent scores each CV against the role rubric, writes a structured summary with evidence for each dimension, and ranks the full applicant pool. Recruiters open their dashboard to a sorted, annotated shortlist.

04

Interview booking automated

Top-ranked candidates receive an automated email with a Calendly link. Confirmed slots written back into Greenhouse and added to the recruiter's calendar, zero manual coordination.

/ RESULTS

Recruiters spend time on people, not on reading PDFs.

-82%

Screening time per role

Average time from application close to shortlist delivery dropped from two days to under four hours.

Candidates reviewed per recruiter

Same team, same hours, four times the roles filled per quarter because the screening bottleneck was removed.

0

Missed applications

Every submitted CV scored and ranked. No application lost in an inbox or skipped because the queue was too long.

11

Days to production

From kick-off to a live agent processing real applications in eleven days.

/ STACK

The smallest stack that solved the problem.

n8nOpenAIGreenhouseCalendly
/ YOUR SYSTEM

An agent that handles
your repetitive process.

30-minute call. We'll map the highest-volume manual step in your workflow and scope the automation, free, no deck.