Recruitment stopped working
when applications became free.
AHRAI reads what a CV actually evidences — transferable skills, behavioural capability, workforce readiness — and gives both sides of the labour market the same intelligence. Watch how, then use the live platform.
Screen the pile, keep the people
Every applicant ranked with the reasoning attached, segmented into hire-ready, developable and stretch potential, with a retention likelihood before you interview anyone.
Know the gap before you apply
An employability score against any role, the exact skills holding it back, and a costed pathway to close each one.
Intelligence, not filtering
Skills supply against demand across a whole vacancy book — workforce planning an applicant tracking system structurally cannot produce.
Every available candidate, searchable.
The pool persists between vacancies. Search by skill, filter by segment, and add your own CVs — PDF, Word or text — to have them scored on arrival.
Find out how ready you are — before you apply.
Paste a CV and choose a target role. AHRAI removes personal identifiers, extracts skills and behavioural evidence, and returns an employability score with the gaps that are holding it back.
Your workforce data
No analysis yet
Load the sample CV, tick consent, and run the engine to see scoring, skills gaps and development pathways.
Screen a full applicant pool in seconds, with the reasoning attached.
Choose a live vacancy. AHRAI scores every applicant on skills, transferable skills, behavioural evidence, progression and readiness — then ranks them and shows you exactly why.
What your applicant market actually contains.
Every applicant scored against every open vacancy — supply against demand, segment mix, pipeline economics and the skills the UK market is not supplying you.
The model changes when hiring outcomes come back.
Record what actually happened to a candidate and the engine re-fits its own weights, versions itself, and reports whether its separation between successful and unsuccessful hires improved.
Five revenue streams, one platform.
Pricing as modelled in the three-year forecast, with a live return calculation against a customer's own recruitment volume.
The engine, opened up.
Every score in this demo comes from the weighted model below. Adjust a weight and the whole platform re-scores — nothing is hard-coded or pre-baked.
Scoring model weights
How a score is produced
Integration surface
Fairness & compliance safeguards
- PII redaction before scoring — names, emails, phone numbers, postcodes, dates of birth and nationality are stripped from the text the model sees.
- No protected characteristics as features — the model has no input for age, sex, ethnicity, nationality or disability.
- Human in the loop — AHRAI ranks and explains; it never auto-rejects. Every shortlist requires a recruiter action.
- Full explainability — each score decomposes into named, inspectable signals with the evidence that triggered them.
- Data minimisation — this demo runs entirely client-side; no CV text leaves the browser.
- Audit trail — production build logs model version, weights and evidence per decision for ICO and EU AI Act Art. 14 review.
MVP today → production model
Identity-invariance test
The same CV is scored twice: once as submitted, once with the name, nationality and postcode replaced. If any identity signal reached the model, the two scores would differ.