AI HR & Recruitment Agent Development

Description

We hire about 200 people a year and receive roughly 14,000 applications doing it. Our recruiters read the first 60 CVs for a role properly and skim the rest, which is an honest description of the problem rather than a criticism of them. We want help with structure, not judgement. Extract skills, experience and availability into comparable fields so a recruiter can sort and filter, and summarise each CV against the role's stated requirements with the evidence quoted from the CV itself. No scoring, no ranking, no automatic rejection. Our legal team has been clear and we agree with them: a system that filters candidates out without a human is a discrimination claim waiting to be filed, and we would not be able to explain how it decided.

• Extract structured fields from CVs in our top formats, including the PDF exports that are one big image
• Summarise each application against the role requirements, quoting the supporting text from the CV
• Explicitly avoid scoring, ranking or auto-rejecting — surface, never decide
• Strip name, age, gender and photo from the recruiter-facing summary view by default
• Log extraction confidence so recruiters know when to read the original

• CV extraction pipeline integrated with our ATS
• Requirement-matched summaries with quoted evidence
• Anonymised default review view
• Extraction confidence reporting and a fallback-to-original flow