SCORED, RANKED, HUMAN-DECIDED
Sorted by AI. Decided by you.
AI CV screening can extract application details and rank them against an employer-defined rubric to help prioritise review. In this proposed scope, a person reads each CV and makes every shortlisting, hiring or rejection decision. The employer reviews the criteria for job relevance and discrimination risk before use; a neutral-looking rubric can still be biased.
Updated
What you’ll find here.
States plainly what AI CV screening can and cannot do under a bias-risk, human-review-required framing, with a full worked scoring rubric shown rather than described, aimed at any small business hiring, not a recruitment agency's own client work.
- The proposed human-review policy is explained alongside ICO guidance on automated decisions; it is not a blanket statement of the law
- The bias-risk framing is cited to the Equality Act 2010's definition of indirect discrimination
- The scoring rubric is shown in full, with its points and its threshold, not just claimed
The job, broken down.
- Reading and extracting details from every CV received
- Scoring each CV against a rubric built from the job description
- Ranking candidates for a person to read
- Presenting every proposed rejection for individual human review
- Flagging anything the rubric is unsure about
The same job, handled.
| The task today | With AI doing the legwork |
|---|---|
| Every CV is read from scratch, in whatever order it arrived | CVs are scored and ranked against the job's own requirements first |
| A busy week means CVs pile up unread for days | Applications can be ranked for review after the agreed checks have completed |
| A rubric's bias, if any, stays invisible | Every proposed rejection requires a person to review the CV; additional sampling can help assess rubric quality |
Built in steps.
Connect
We connect the inbox or applicant tracking system applications already arrive in, so nothing needs re-entering by hand.
Extract and score
Each CV is read and scored against a rubric the employer defines in advance, built from the job's actual requirements rather than a generic profile.
Rank for review
Candidates are ranked to help prioritise review. Every proposed rejection is surfaced for an individual human decision; quality sampling never substitutes for that review.
Hand over
A person reads the shortlist and makes every shortlisting and hiring decision. Nobody is auto-rejected without a human looking at the CV.
Do the maths.
Reading and recording application details for recurring recruitment by hand.
| Input or step | Value |
|---|---|
| Times a week | 50 |
| Minutes each time | 5 |
| People doing it | 1 |
| Cost per hour | £18 |
| Hours a year | 217 (about 29 working days) |
| Cost a year | £3,900 |
Illustrative weekly workload: fifty applications a week, five minutes each to read and score manually. The annual figure assumes this workload for all 52 weeks; it is not the cost of a single vacancy or an AI saving. Use your actual weekly average, including weeks without hiring. See how we calculate this.
These are illustrative workflow scopes. The audit checks the actual software, supported access, licences, permissions and human review rules before a build is agreed.
What AI CV screening actually does
A small business hiring for one role rarely has a recruitment team behind it, just whoever is already busy running the business, now also reading fifty CVs. AI screening reads each application, pulls out the details that matter for that specific role, years of relevant experience, named qualifications, particular skills mentioned, and scores it against a rubric built from the job description. The output is a ranked shortlist for a person to read properly, not a decision.
A worked scoring rubric, shown in full
This is illustrative, built for one example vacancy, not a real job’s actual criteria:
| Criterion | Points available |
|---|---|
| 3 or more years in a directly relevant role | 3 |
| A named, relevant qualification | 2 |
| Evidence of two named core skills from the job description | 2 (1 each) |
| Maximum score | 7 |
An illustrative threshold of 4 can prioritise applications for review, but it cannot reject anyone. Under this proposed scope, a person reads each CV before any rejection or shortlisting decision. Additional sampling may help assess the rubric, but cannot replace individual review. The criteria and threshold require employer review for job necessity and discrimination risk.
Where the bias risk actually sits, and who carries it
Section 19 of the Equality Act 2010 defines indirect discrimination as a practice that applies to everyone but puts people sharing a protected characteristic at a particular disadvantage, unless it can be shown to be “a proportionate means of achieving a legitimate aim.” A rubric that scores continuous, unbroken employment highly can do exactly that, disadvantaging someone who took time out for a disability, for parental leave, or for caring responsibilities, without ever mentioning any of those things. That is why a rubric is built from what the job genuinely needs, checked before it is used, and is reviewed for criteria that could act as proxies for protected characteristics; a neutral-looking rubric does not guarantee absence of bias.
The ICO’s guidance on Article 22 of the UK GDPR names “a decision to offer someone a job” directly as an example of a decision with a similarly significant effect on a person, a reason to assess automated-decision risk carefully. That guidance discusses exceptions and carries a data-law review notice; this page does not claim a blanket statutory prohibition. The proposed workflow policy is stricter: a person reads the actual CV and makes every shortlisting, hiring and rejection decision.
On the tools you already use.
- Your applicant tracking system or inbox
- Job description and person specification
- Google Sheets or Excel for the shortlist
Who needs it most.
What it costs.
Every price is published and fixed in writing before any work starts.
AI Opportunity Audit
In person across Leeds and West Yorkshire, or by video call. A written report of your top 5 automations, yours to keep.
AI Lead & Ops Sprint
Enquiry capture through to booking, follow-up and CRM logging, plus one ops automation. The monthly fee covers monitoring, tuning, a monthly review and up to 2 hours of changes.
Fractional AI Team
Ongoing AI builds, staff training and a monthly review.
Prices exclude VAT where applicable.
The build guarantee. We agree written acceptance criteria before any work starts. If the build doesn’t meet them, we fix it at no extra cost, and if we can’t, you don’t pay for it.
Where the facts come from.
- What is the impact of Article 22 of the UK GDPR on fairness?Information Commissioner's Office
- Equality Act 2010, Section 19: Indirect discriminationlegislation.gov.uk
Questions, answered straight.
Can AI reject a candidate on its own?
No. AI CV screening set up by Neura North scores and ranks candidates against a rubric; it does not reject or hire anyone. This is the proposed workflow policy, not a blanket statement of automated-decision law. The employer must assess the actual use and relevant data-protection and discrimination requirements.
Can a CV scoring rubric be discriminatory even if it never mentions protected characteristics?
Yes. Section 19 of the Equality Act 2010 defines indirect discrimination as a practice that applies to everyone but puts people who share a protected characteristic at a particular disadvantage, for example scoring against a continuous employment history in a way that disadvantages people who took parental or disability-related leave. Every rubric is checked for this before use.
What does the scoring rubric actually look like?
The illustrative rubric uses stated job requirements, qualifications and skills to prioritise human review. The employer checks whether each criterion is necessary and could disadvantage a protected group; excluding names and ages does not guarantee a bias-free result. Every application still receives human review before a decision.
Is this the same as Neura North's recruitment agency page?
No. This page is for any small business screening its own applicants for its own vacancies. The recruitment page covers a recruitment agency automating its client-facing placement work, a different job entirely.
What does AI CV screening cost?
A workflow like this is usually a fixed-price package agreed in writing before any work starts (most clients begin with the AI Opportunity Audit at £495, credited against a build). The Fractional AI Team at £1,500 a month covers ongoing work if you want it looked after. Prices exclude VAT where applicable.
Where to go next.
Ranked by a rubric. Chosen by someone.
Start with a 20-minute call.
Tell us the job that keeps coming back. We’ll tell you honestly what’s worth automating and what isn’t, and put a number on each one.