SanRa
Charcoal study of David casting a sling; the stone rendered as a polished AI pebble.

CREST Certified in Cybersecurity Training

SanRa

Discernment in the age of machines.

We help leadership teams govern AI and cybersecurity with the same rigour they apply to finance and operations — combining specialist advisory with software built for the decisions that matter.

Request a proposal How we engage

01 / 04

01 — The challenge

Capability without
judgement is risk.

Most organisations already use AI — in products, procurement, support and shadow tooling — while ownership, evidence and escalation paths lag behind. The failure mode is rarely the model alone; it is the gap between what systems can do and what people are prepared to decide.

  • Unmapped AI use across the enterprise
  • Over-reliance on automated outputs
  • Accountability that stops short of the board

02 — Human factors

Controls work only
when people can use them.

We study the real conditions of work: incentives, time pressure, handovers and trust in machines. That is where otherwise sound security programmes erode — and where durable improvement begins.

  • Behaviour, culture and usable security
  • Social engineering and influence risk
  • Decision quality when systems advise

03 — AI security

Secure the system,
not only the model.

We treat AI as a sociotechnical system: data, models, integrations, operators and suppliers. Threat modelling and assurance follow the full lifecycle — from design and procurement through to production and retirement.

  • Lifecycle threat modelling
  • Data, model and pipeline protection
  • Adversarial testing and assurance

04 — Build & govern

Advice that can
be implemented.

Where frameworks need operating reality, we design the governance and — when needed — the software. ISO/IEC 42001 readiness, control evidence, and tailored tooling sit in one coherent engagement.

  • ISO/IEC 42001 readiness & AIMS design
  • Custom security and assurance software
  • Evidence packs for leadership and audit

Engagement model

From mandate
to measurable outcome.

Structured like a professional services engagement: clear scope, accountable delivery, and outputs your leadership can act on — including advisory and bespoke software where required.

AI security self-assessment

Where do you stand?

Three questions. A useful signal — not a formal audit.

01 Do you know where AI is used across your organisation?
02 Are AI systems assessed for threats before deployment?
03 Is there clear accountability for AI risk?

Indicative only. Not a certification, audit, or compliance determination.

Whitepaper

ISO/IEC 42001 readiness

A concise guide to the decisions and evidence behind a credible AI management system.

Download the guide

Credentials

CREST Certified in Cybersecurity Training

Consultant(s) certified by EC-Council and Proofpoint. Guest lecture at SANS. Partners: Erasys Ltd. and TAAM Open Source.

Consultation

Discuss a mandate.

Share your context and constraints. We respond with next steps — typically within one business day.

hello@sanra.co.uk