Skip to content
AI Vector

Trust & security

An assessment is an act of trust. We hold up our end.

You are telling us where your organization is strong and where it is exposed. These commitments are structural, not aspirational.

Your data trains nothing

Assessment answers and reports are never used to train AI models - ours or anyone else’s. AI providers we use are contractually barred from training on your data.

Reproducible scores

Every score is computed deterministically from your stored answers by versioned rules. We can re-derive any result you ever received, and so can you.

Export and deletion, self-service

Download everything you have given us, or delete it, from your account - no support ticket, no retention tricks.

Benchmarks without leakage

Peer comparisons activate only at 30+ comparable organizations, computed on aggregates. No customer ever sees another customer’s data.

AI with a leash

The AI interviewer explains and clarifies; it never sets scores. Its outputs pass groundedness checks, and a deterministic fallback exists for every AI-powered step.

Boring, proven infrastructure

EU-headquartered hosting provider, encrypted in transit and at rest, nightly backups with offsite copies, and access limited to what operating the service requires.

AI Vector is operated by A.G. Logik, Connecticut, USA ([email protected]) and is offered to customers in North America, Central America, South America, and the Caribbean Islands. Purchases are billed in US dollars and carry a 14-day full refund policy. See our Terms of Service and Privacy Policy, which lists every subprocessor that touches your data.