AI for recruitment
A job ad draws a hundred and twenty applications in ten days. Thirty are read, ten called, one hired — and a hundred and ten people never hear back. The task that costs in recruitment is not the choice, it is everything around it: acknowledging, replying, booking interviews, chasing a hiring manager, writing up notes. Half a day, every day, in which the recruiter does not recruit.
The tasks we remove first in this industry
What this industry pays when nothing moves.
Ranges, not promises. They come from tasks we meet in this industry and orders of magnitude we measure in the survey — yours may fall outside, in either direction.
What we remove, and what we leave.
Each task is described with what the agent takes on and what stays with a person. If the second column is empty, we have not understood the industry.
Acknowledge and answer every application
What the agent does
Every candidate gets an acknowledgement within the hour, then a follow-up — interview offered or not — written in your tone. Never silence.
What stays with a person
The choice. The agent does not sort, rank or score: the recruiter reads and decides, and the log shows no one was screened out by a machine.
Draft the job ad from the brief
What the agent does
From the hiring manager’s notes and previous ads, it drafts the role description and the advert, with no discriminatory criteria.
What stays with a person
Review and publication. An unlawful line in an advert commits the employer.
Schedule interviews
What the agent does
It offers slots from the recruiter’s and manager’s calendars, confirms, and sends reminders and preparation to the candidate.
What stays with a person
Arbitrating between two candidates for one slot, and everything said in the room.
Chase candidates and managers
What the agent does
A manager who has not given feedback in four days, a candidate who has not confirmed: it chases at the right moment, in the right tone.
What stays with a person
The content of the feedback. It asks; it does not suggest.
Write up interview notes
What the agent does
From the recruiter’s notes or transcript, it produces the write-up in your template: what was said, points to check, next steps.
What stays with a person
The assessment. The write-up reports facts; the view on the person is written and signed by the recruiter.
No candidate is rejected, ranked or scored by an agent. Ever.
Machine screening of people is the most direct discrimination risk there is, and the law frames it. The agent replies, schedules, chases and writes up; it does not read a CV to decide. The recruiter reads every file, and the log proves each decision was human. The diagram below shows where its reach ends.
No branch contains a filter. The condition is about what the recruiter decided — never about the candidate.
Ten systems, one survey, one system opened at a time — the one with the shortest payback. Read the full method, or go back to the AI agency page.
What we open it with, and why here.
The choice of tool comes last, and it is reversible. But in this industry, three constraints keep coming back and steer the choice before we even start.
The agent reads what your tool exposes and writes where it allows. It replaces nothing — it finally makes it reply.
Every reply and reminder is timestamped with its cause. If a candidate challenges, the log shows a recruiter read the file and decided.
It writes in your house tone from past replies. As soon as a task calls for assessing a person, it hands over.
What we hear in this industry, before we start.
We get three hundred applications per role. We have to filter.
Yes — and the recruiter filters, against written criteria, on every file. The agent presents files in arrival order, schedules and replies. What it removes is the two hours of email and calendar a day, not the reading.
How long before it is useful?
The survey of the ten systems is delivered within 72 hours. Opening the first system takes two to six weeks. In a talent team, the reply to every candidate often goes live in the first week.
And for an agency rather than an in-house team?
The tasks are the same, plus a client to keep informed: the agent prepares the weekly client update from the ATS. Choosing and presenting candidates stays with the consultant.
The areas where we are called most for this.
The same tasks, in neighbouring industries : AI for HR, AI for sales teams, AI for legal teams.
We will tell you which system to open first.
Describe the task that costs you the most. We will tell you what is feasible, how long it takes and what it costs. If the answer is no, you will leave with the two reasons why.