CV ranking your Data Protection Officer can approve.
AI-assisted CV screening in a sandboxed environment — candidate PII tokenised before reaching the model, justification logged for every ranking decision, and sensitive data never leaving your workstation. GDPR-compliant AI recruitment that is actually faster.
Three compliance risks in every AI-assisted recruitment workflow.
Speedwave resolves each at the architecture level — not through process workarounds.
GDPR and candidate PII
Feeding CVs into ChatGPT or Copilot means candidate names, addresses, phone numbers, and national IDs are processed by a third-party model with no specific data processing agreement. That is a GDPR Art. 28 risk every time you screen a candidate.
Bias and explainability requirements
Regulators and internal compliance teams increasingly require documented justification for AI-assisted shortlist decisions. "The AI ranked them lower" is not a defensible answer under GDPR Art. 22 (automated decision-making).
Manual screening overhead
Screening 200 CVs against a competency framework is time that could be spent in candidate conversations. The pattern-matching work — certifications, years of experience, domain exposure — is automatable with the right safeguards.
What Speedwave gives HR and recruiters.
AI-assisted recruitment that is fast, auditable, and GDPR-compliant out of the box.
PII tokenisation
Candidate names, email addresses, phone numbers, national identity numbers, and addresses are tokenised before CV content reaches the model. Tokens are stable within a session — ranking logic works correctly — but real personal data never leaves your workstation.
Documented ranking justification
Every shortlist ranking includes a structured justification: which competencies were matched, which were absent, and at what confidence level. The log is review-ready for your compliance team or an external audit.
Natural language screening
Describe the ideal candidate in plain English — required certifications, domain experience, years of relevant work. Speedwave screens against that description and returns a ranked shortlist with per-candidate justification.
On-workstation processing
CV data is processed in a Speedwave container on the recruiter's workstation. No CV content passes through Speednet infrastructure. The data processing boundary is your machine — not a vendor cloud.
Audit trail for shortlist decisions
Every screening session, ranking output, and recruiter override logged with timestamp. Useful for GDPR Art. 22 documentation, internal compliance review, or if a candidate requests an explanation of why they were not shortlisted.
Ready to see it in your BFSI environment?
We will walk through a live CV screening demo — including PII tokenisation and the audit log output — in 30 minutes.