Screen

First-round screening calls that actually happen.

An AI voice agent runs the first screening conversation, then the call becomes a structured report a human can act on — availability, notice period, expectations, and the answers to the questions you set.

Who it is for
Teams whose shortlist decays while first calls wait to be scheduled
The problem
Fifty shortlisted candidates and one recruiter means the fiftieth first call happens three weeks late, by which point the candidate has taken another offer.

What it does

The short version.

  • Consent disclosure that cannot be turned offEvery call opens by saying what it is.
  • A hard cap on attemptsNobody is called until they stop answering.
  • Calls become structured reportsA readable summary with the answers, not a recording to sit through.

In detail

What is actually in here.

The modules behind this stage. Each one is specified and built against a written brief before it ships.

  • AI screening callsOutbound voice screening through the Bolna adapter, with a call script, retry policy and attempt cap.
  • Screening reportsThe call summarized into a structured, reviewable report attached to the application.

The AI boundary

What the model offers, and what you own.

AI proposes

The conversation and the summary of what the candidate said.

A person confirms

A recruiter reads the report and decides. The call is disconnected by default and needs explicit confirmation in the UI before it can dial anyone.

Try it on one role.

Create a workspace and open a job. Nothing here requires you to migrate anything first.