Does AI decide who gets hired?

No. It reads applications, checks them against criteria you published, and writes down why. Every decision about a person is made by a person, and the rejections are approved by a human before they send.

Can it handle CV attachments?

Yes. The attachment is parsed into the fields your workflow reads, so screening happens on structured data rather than on whoever has the patience to open forty PDFs.

Should rejected candidates get a reply?

Yes, and quickly. A kind no within days beats six weeks of silence, and it is the cheapest reputation work a company can do. Today's decline is next year's strong hire or referral.

Can it sort applications for several roles at once?

Yes — the role is just another branch. Each posting has its own criteria and its own queue.

Is automated CV screening legal?

Broadly, with obligations attached, and they are growing. New York City's Local Law 144 requires bias audits and candidate notice for automated employment decision tools, and the EU AI Act treats hiring systems as high-risk with human oversight required. Written criteria, human-owned rejections and a decision log are what those rules look like in practice. Ask your employment lawyer about your situation.

← All use cases

jobs@

Recruiting email automation

One decent job ad brings in a few hundred applications, and the honest reason most of them never get a reply is that answering them by hand is a full-time job. An agent on jobs@ answers all of them and sorts the pile. A recruiter still decides who moves forward.

What it looks like today

Applicants who hear nothing for three weeks and assume the worst, correctly.
The strong candidate who was in the pile the whole time and accepted somewhere else on day nine.
Three emails and two days spent agreeing on an interview slot.

How the jobs@ workflow runs

1
Every application gets acknowledged
Within minutes, from a template, identical for everyone: we have it, here is what happens next, here is roughly when you will hear back. No AI involved — this message should not vary by candidate.
2
The application becomes fields
Current role, relevant years, location, the skills the ad asked for, portfolio links. Sorting works on fields. Skim-reading PDFs at seven in the evening does not.
3
It gets checked against what you published
The criteria in the job ad, and nothing else. What comes out is an assessment with reasons attached, not a score with no working shown.
4
Three branches, and a person owns two of them
Clear match: a draft offering interview times, queued for the recruiter. Borderline: a review queue with the assessment attached. Clear miss against the published criteria: a kind decline that a recruiter still approves.

What changes

Nobody is left guessing. Every applicant gets a real answer within days.
Your recruiter starts from a sorted queue with reasons attached instead of an inbox.
The scheduling back-and-forth stops eating afternoons.

Where you stay in charge

This is the mailbox where the guardrails matter most, so they are not optional here.

Screening runs against the requirements you published for the role. If a rule is not written down where the hiring team can read it, it is not a rule.
A person owns every rejection. The workflow gathers the evidence and drafts the message; a recruiter reads it, decides, and puts their name on it.
Keep the inputs that quietly stand in for protected characteristics out of it — graduation year, commute distance, gaps in employment. A useful test: could you explain this requirement to the candidate it excluded?
Every classification, draft and approval is logged. 'The software decided' is not an answer you want to give a candidate, and it is not one a regulator accepts either.

One morning, concretely

Fifty applications arrive overnight. By nine every one of them has been acknowledged, parsed and assessed against the ad, and the recruiter opens a shortlist with reasons attached instead of a raw inbox.

Questions people ask

Does AI decide who gets hired?: No. It reads applications, checks them against criteria you published, and writes down why. Every decision about a person is made by a person, and the rejections are approved by a human before they send.
Can it handle CV attachments?: Yes. The attachment is parsed into the fields your workflow reads, so screening happens on structured data rather than on whoever has the patience to open forty PDFs.
Should rejected candidates get a reply?: Yes, and quickly. A kind no within days beats six weeks of silence, and it is the cheapest reputation work a company can do. Today's decline is next year's strong hire or referral.
Can it sort applications for several roles at once?: Yes — the role is just another branch. Each posting has its own criteria and its own queue.
Is automated CV screening legal?: Broadly, with obligations attached, and they are growing. New York City's Local Law 144 requires bias audits and candidate notice for automated employment decision tools, and the EU AI Act treats hiring systems as high-risk with human oversight required. Written criteria, human-owned rejections and a decision log are what those rules look like in practice. Ask your employment lawyer about your situation.