About the founder
CIPD Level 7 qualified HR professional. A decade in workforce compliance, training, and applied research. Mermoid is what happens when an HR practitioner—not an engineer playing recruiter—builds the AI scoring tool the sector actually needs.
Every AI hiring tool I evaluated as a CIPD-qualified practitioner had the same problem: it produced a score, but it could not defend the score. No citations. No audit trail. No way to explain to a candidate, a tribunal, or the ICO why the model said what it said.
Under the UK Data (Use and Access) Act 2025 and the ICO's April 2026 ADM guidance, “decision support” that recruiters rubber-stamp is automated decision-making in disguise. Under the EU AI Act, employment-context AI is high-risk by definition. The current generation of LLM-only scorers is not defensible under either.
Mermoid is the deterministic alternative: every score has a citation back to a CV field and a JD field. The math is reproducible. The provenance is durable. It's how an HR practitioner would build an AI tool, because that's what I am.
Independent recruitment consultancy serving CQC-registered care providers, FCA-regulated firms, SRA-regulated practices, and other regulated sectors. Operator of the Mermoid audit-first scoring platform.
A decade of front-line workforce compliance: training design and delivery, CQC Reg 19 audit-trail maintenance, multi-agency liaison, person-centred care planning, and reablement support. Hands-on familiarity with the documents, audits, and inspections recruiters need to clear before placing a candidate.
Curriculum-aligned teaching; advanced research tools (SPSS, NVivo) applied in teaching and research contexts.
Civil proceedings; case documentation for judges. Procedural literacy across order codes and cross-departmental process.
Therapeutic sessions for resilience-building; safeguarding liaison; multi-agency planning.
My research background (PGDip Social Research Methods) is the uncomfortable truth behind Mermoid's design. AI scoring is research: a hypothesis (this candidate fits this role) tested against operationalised criteria. If you can't reproduce the score, you don't have a finding—you have a guess.
Published research includes a journal article on Lincolnshire County Council's Troubled Families Programme (developed with the council's Early Help Teams; held on Academia.edu and ResearchGate) and a payday-loan study that contributed to a University of Lincoln policy change.