Ten to two hundred people, and every résumé lands in your inbox between other work.
Hiring was added to your job; the ATS, the inbox and the job boards never agree.
Hundreds of applications per role, many of them AI-written, and interviews to schedule by hand.
Restaurants, clinics, logistics, retail and home services, where speed to fill is revenue.
Numbers from Gem, SHRM and recruiter surveys — the volume has tripled while teams shrank, and most of the hours go into screening and scheduling.
applications per recruiter a year in 2024, against 925 in 2021 — 2.7 times more, with 20% fewer recruiters.
Gem 2025 Recruiting Benchmarks ↗of applications end in a hire, down from 1.6% in 2021 — most screening time goes to candidates who will never be hired.
Gem 2025 Recruiting Benchmarks ↗average time to hire in 2024, up from 33 in 2021; interviews per hire rose from 14 to 20.
Gem 2025 Recruiting Benchmarks ↗of recruiters say scheduling a single interview takes 30 minutes to 2 hours.
Recruiter survey cited by Kula ↗year-over-year growth in LinkedIn applications — hiring is becoming a machine-versus-machine contest.
eMarketer ↗of employers with hiring difficulties cite candidate ghosting; 50% cite competition from other employers.
SHRM Talent Trends 2025 ↗Put your monthly volume into the calculator and see hours and dollars.
Recruiting automation replaces the manual chain behind every opening — gather applications from the ATS, the inbox and three job boards, read each résumé, send the same pre-screen questions, trade emails to find an interview slot, chase interviewers for feedback, remember to send rejections, retype the hire into payroll — with a workflow that does the routine steps on its own. Applications per recruiter have nearly tripled in three years, and most of the hours now go into screening people who will never be hired and into scheduling. That is the part automation in recruiting takes over, while every decision about a candidate stays with a person.
We build it around the ATS and calendars you already use. A new application in Greenhouse, Lever, Workable or BambooHR is parsed and scored by a language model against the rubric you wrote, with a written reason in the candidate card; strong profiles receive pre-screen questions by email or SMS, and after the recruiter’s approval the candidate self-books an interview against the interviewers’ real calendars. Scorecards are chased in Slack, feedback is summarised for the debrief, rejections go out within a day and a signed offer creates the employee in the HRIS without a second data entry.
Candidates hear back in hours instead of weeks, interviews are booked without a single email thread, no one is left without an answer and recruiters spend their time on conversations rather than on sorting. Wireclad builds this recruitment automation with n8n, APIs and language models on your own infrastructure, with logs that make bias checks and audits possible — automation for recruiters, not instead of them.
The run path of a typical recruiting workflow. Each application travels it on its own; every decision about a person is made by a person.
A Greenhouse, Lever, Workable or BambooHR webhook fires; applications arriving by email or from job boards are parsed into the same run.
Extracts experience, skills, locations and dates from the PDF into a structured profile.
A language model compares the profile with the criteria you wrote for the role and writes a score with a two-line reason into the candidate card.
Strong profiles move to pre-screen; borderline ones go to the recruiter’s review list; nobody is rejected by the model.
Two to five role questions by email or SMS through Gmail and Twilio; the answers return to the ATS.
After the recruiter’s thumbs-up, the candidate gets a self-booking link with the interviewers’ free slots from Google Calendar or Outlook.
Books the slot, creates the Zoom or Meet link, reminds both sides and re-offers slots if someone cancels.
When the calendar event ends, the interviewer gets the scorecard in Slack or Teams; escalates to the hiring manager after 24 hours.
Collects the scorecards into a short comparison for the debrief — the decision is made by people.
A status change to “Rejected” sends a personalised, polite email within a day — nobody is left without an answer.
Fills the offer template, sends it via DocuSign and starts the Checkr check in parallel.
A signed offer creates the employee in BambooHR, Gusto or Rippling and starts onboarding — no second data entry.
Sources, time to hire, stage conversion and open scorecards in Google Sheets or Looker Studio, summarised in Slack.
Typical minutes per application for a recruiter working by hand. Your own numbers go into the calculator below.
Based on 16 min by hand and 1.1 min with the flow per item, from the table above.
The flow collects, scores, schedules and reminds. People keep every decision about a candidate.
The model scores and explains; a recruiter and the hiring manager decide.
Conversations, culture fit and judgement stay with people.
Approved by a person before the template is filled.
Reviewed by a recruiter, also to check the model for bias.
Anything off-script goes to the recruiter, not to a template.
EEOC rules and audits such as NYC Local Law 144 are owned by you; the flow keeps the logs that make them possible.
Flip a switch to hand a step to the flow or take it back.
Benchmarks from Gem’s 2025 Recruiting Benchmarks, SHRM Talent Trends 2025 and a recruiter survey cited by Kula: the volume has nearly tripled while teams shrank — the screening and scheduling load is what a workflow absorbs.
We map the process as it runs today, count the minutes and agree what the flow must never do on its own.
A working flow on your real data, in a sandbox. You see every run and every exception.
Edge cases, approvals and alerts, then the switch-over — with the old way kept as a fallback.
Monitoring, fixes when a vendor changes a format, and a monthly report of hours saved.
Yes. The flow listens to Greenhouse, Lever, Workable or BambooHR webhooks, writes the score and reason back into the candidate card and moves stages through the API. Your recruiters keep working where they already are.
The model scores only against the job criteria you write, never rejects on its own, and every score comes with a logged reason. Borderline and rejected candidates are reviewed by a person. Where rules like NYC Local Law 144 or EEOC guidance apply, the logs make the audit possible — and you remain the decision-maker.
Recruiters report 30 minutes to 2 hours per interview by hand. Self-booking against real calendars with automatic reminders and rescheduling reduces it to a glance at the calendar. Put your own volume into the calculator above.
Yes — pre-screen messages and booking links are clearly from your company, say what happens next and carry a recruiter’s name. Most candidates prefer a fast, clear process to silence; questions outside the script go to a person.
Possibly not the whole pipeline. Scheduling, scorecard reminders and rejection emails pay off even at low volume because they remove the dropped balls; full scoring is worth it when applications per role run into the hundreds.
In your ATS. The workflow runs on your server or in your own cloud account, keeps only what a run needs, and the scoring model can be a provider that does not train on your data or a local model. Access and retention follow your policies.
Then applications from email and job boards are collected into a simple pipeline in Airtable or Google Sheets with the same scoring, pre-screen and scheduling steps, and you can move to an ATS later without losing the process.
Describe it in two sentences — we reply within a day with a workflow sketch.
hello@wireclad.com →