CV Screening Automation: Every Application Read, Scored and Shortlisted, With a Reason for Each

Why good candidates get missed

CV screening automation fixes the part of hiring every owner dreads: one job post, two hundred applications. They arrive by email, in a job portal, as PDFs on WhatsApp and as forwarded messages from friends. Somebody has to open each one, read it, and decide. That somebody is usually the owner or a manager who already has a full day.

So the reading gets rushed. The first thirty CVs get a careful look, the rest get a glance late at night, and the best candidate may be sitting at number 140, unread. Good people take another offer while you are still working through the pile. The candidates you say no to hear nothing at all, which is not how you want your business remembered.

This system reads every application the same way, against the must-haves you set for the job. Each CV gets a score and a short reason. The top ten land in one sheet, ready for you to review, and every candidate gets a polite update. You still make every hiring decision; you just start from a shortlist instead of a pile.

What gets automated

• Every application in one place

  CVs from Gmail or Outlook, job portal exports and WhatsApp are collected into one folder and one sheet, so nothing sits in a forgotten inbox.

  • Email attachments
  • Portal exports
  • WhatsApp PDFs

• Read against your must-haves

  An AI model reads each CV against the job's must-haves and nice-to-haves that you write once, such as experience, skills, location and notice period.

  • Your criteria only
  • Same rules for all
  • No skipped CVs

• A score and a reason

  Every candidate gets a score and two or three lines on why, so you can see how each decision was made and overrule it.

  • Score per CV
  • Reason in plain words
  • Missing details flagged

• A top 10 shortlist

  The strongest candidates land at the top of one Google Sheets or Excel list with their contact details and the CV link.

  • Ranked list
  • CV linked
  • Ready to call

• Polite updates to candidates

  Every applicant gets a short message on email or WhatsApp: received, shortlisted, or not this time, with a thank-you.

  • Received note
  • Next steps
  • Kind no

• Unsure CVs to a person

  Applications the AI cannot judge, such as a career change or an unreadable file, go to a person instead of being dropped.

  • Human review
  • No auto rejection
  • Notes attached

How the workflow runs

Who it is for

What you need to start

You need the places applications already arrive: a job email address in Gmail or Outlook, exports from the job portals you use, and a WhatsApp Business API number through a provider such as WATI or Twilio if candidates send CVs on WhatsApp. The shortlist lives in Google Sheets or Excel and the CVs in Google Drive or OneDrive. The workflow is built on n8n, Make or Zapier, with ChatGPT or Claude doing the reading.

The most important input is yours: a clear list of must-haves and nice-to-haves for each role, and the wording of the messages candidates receive. We help you write criteria that are about the job, so the system does not judge people on things that should not matter, and we keep a person in the loop for every final decision. Candidate data is personal data, so access is limited to the people who hire.

A typical build runs over a few weeks: it is set up for one live role first, the scores are checked against your own judgement on real CVs, and the criteria are tuned before it is used for the next hire.

Questions about CV screening automation