RexaThe short answer
A recruiting voice agent calls every applicant within minutes of application, asks the same qualifying questions in the same order, and writes structured answers — availability, notice period, salary expectation, location, work authorisation — back to your system. Recruiters then spend their time on shortlisted candidates rather than on voicemail. Calls that need a human transfer live to a recruiter on request.
Last reviewed
What does a recruiting voice agent actually ask?
You define the script. A typical first-round screen confirms the candidate is still interested, checks availability and notice period, asks the two or three knockout questions for the role, confirms location and work authorisation, and offers interview slots. Because every candidate hears the same questions in the same order, the answers are directly comparable — which is difficult to achieve across a team of human screeners working from a shared document.
How fast can it call a new applicant?
Calls can be triggered the moment an application lands, within your configured calling windows. Speed to first contact is the variable recruiters most often cite as decisive, and it is the one an automated caller changes most directly — an agent does not have a queue, a lunch break, or forty other applicants ahead of yours.
What happens when a candidate wants a person?
The agent transfers the live call to an available recruiter, passing the conversation context. If nobody is free, it captures a callback preference and the call is queued with the transcript attached, so the recruiter who picks it up starts informed rather than cold.
Does it work for inbound candidate calls too?
Yes. The same agent answers inbound calls to your recruiting line, identifies the candidate and the role they are calling about, answers common questions about the process, and books or transfers as appropriate. Inbound and outbound run on the same platform at the same per-minute rate.
Frequently asked questions
- Can an AI voice agent screen candidates fairly?
- A voice agent asks every candidate the same questions in the same order, which removes the ordering and fatigue variation that affects human screening. It does not remove bias from the questions themselves — that remains your responsibility. Review your script for job-relevance, keep the structured answers auditable, and have humans make the shortlisting decisions.
- Will candidates know they are talking to an AI?
- Yes. Disclosure is configured in the opening line of the script, and in several jurisdictions it is required. We recommend disclosing in every market regardless, because candidates who discover it mid-call react far worse than candidates told up front.
- Does it integrate with our ATS?
- Call outcomes are emitted as structured fields over webhooks, so they can be written into any system that accepts an HTTP callback. See the webhooks documentation for the event payloads and the functions documentation for calling your own endpoints mid-conversation.
Scope, limitations, and sources
This page describes platform capability, not guaranteed outcomes. Results depend on your list quality, script, calling hours, and the regulations that apply to your industry and jurisdiction. Nothing here is legal advice: you remain the caller of record and are responsible for confirming that your campaign complies with the rules in each region you call.
Sources
- Rexa.ai capabilities, per-minute rate, and what the rate includes. rexa.ai (accessed )
- Rexa.ai calling-window, DNC, recording-consent and audit-evidence controls. rexa.ai (accessed )
Spotted something out of date or wrong? Tell us on the contact page and we will correct it and update the review date.