Created a voice AI screening workflow that asks role-specific questions, scores fit, and schedules qualified candidates automatically.
Challenge
A recruiting team was spending too much time on basic screening for high-volume roles. Recruiters chased candidates, repeated the same eligibility questions, manually scored answers, and entered notes into the ATS. Strong candidates were delayed while low-fit applicants still consumed time.
Workflow approach
We built a screening agent that calls or texts candidates after application submission, confirms availability, asks role-specific knockout and scoring questions, records structured answers, and books qualified candidates into recruiter calendars. The ATS is updated automatically with transcript, score, risk flags, and recommended next step.
Observed outcome
Recruiter admin work fell 62%, completed screens increased 3.4x, and candidates received faster follow-up. Recruiters used their time for final evaluation and relationship work instead of first-pass qualification.
Outcome metrics
- Less recruiter admin
- 62%
- More screens completed
- 3.4x
- Saved each week
- 28h
- Candidate completion
- 92%
Systems involved
- GPT-4o
- Twilio Voice
- Greenhouse API
- Calendly
- n8n
- PostgreSQL