CYBER ARMOUR Scoping call
Services

What we are hired to do

Ten services across offensive testing, AI security, and response and assurance. Every one of them scoped against your architecture and your risk.

01Offensive testing

Offensive testing

01

Penetration testing

Objective-driven and scoped around the data and systems that would actually hurt you, across network, application, cloud and specialty environments.

  • Findings tied to business impact, not a list of CVSS scores
  • An independent third-party position for your board, auditor or customer
  • Evidence you can use to justify the next security investment
02

Red team and adversary simulation

Full scope and objective-driven, run against your live defences. From initial access to the crown jewels you nominate, with your defenders unaware it is happening.

  • Find out how far an attacker gets before anyone notices
  • Detection and response measured against real tradecraft, not a checklist
  • A collaborative debrief your blue team learns from
03

Purple team assessment

We work alongside your defenders, running real attack techniques while they watch, so you find out exactly where detection breaks down and why.

  • Layered test cases covering the variations of each technique, not one sample each
  • Knowledge transferred during the engagement, not handed over at the end
  • Repeatable validation your team can run again without us
04

Application security assessment

Web, API, mobile and desktop applications. White-box wherever you can give us source, because reading the code finds what fuzzing the front end never will.

  • Exploitable issues demonstrated against your real data, not theoretical ones
  • Authenticated, multi-role and multi-tenant testing, not just the login page
  • Remediation written for the engineers who have to implement it
02AI security

AI security

05

AI red teaming

Adversaries are using AI to move faster, and your AI systems are themselves a target. We test both: your estate against an attacker working at machine speed, and your models, pipelines and agentic workflows against the attacks that only apply to them.

  • Rehearse the worst case against an adversary that does not get tired
  • See where detection and investigation break down when an attack outpaces your analysts
  • A debrief your team keeps, with the techniques documented well enough to re-run
06

AI security assessment

AI systems fail in ways traditional applications do not. We assess them component by component: the model, the pipeline that feeds it, the infrastructure it runs on, and the agent that acts on its output.

  • Prompt injection, jailbreaks and data exposure tested against your actual deployment
  • Supporting infrastructure assessed from a compromised-user position, because that is how it will be reached
  • Agentic workflows tested for what they can be made to do, not only what they can be made to say
03Response and assurance

Response and assurance

07

Active Directory and identity

Mapping the chains of privilege that turn one compromised laptop into domain admin, across Active Directory and Entra ID.

  • Attack paths prioritised by what they unlock, not by how many there are
  • Choke points identified, so a handful of fixes close many paths at once
  • A measurable reduction you can put in front of an auditor
08

Incident response and forensics

Containment, scope and root cause, in that order. Malware reversed when the answer is in the binary.

  • A defensible timeline of what happened and what was reached
  • The written record your regulator, insurer and board each need, for their own reasons
  • Lessons turned into detections before we leave
09

Security audit and architecture review

Against ISO 27001, NIST CSF and the standards your regulator cares about. We look for the designs that are wrong, not just the settings that drifted.

  • Findings mapped to the framework your auditor will ask about
  • Architectural problems named, not buried in a control matrix
  • A remediation plan costed against the budget and backlog you actually have
10

Social engineering and awareness

Phishing, vishing and pretexting against real targets under agreed rules, followed by training built on what your people actually did.

  • Measured against criteria agreed before the campaign, not chosen afterwards
  • The pretexts that worked documented, so you can defend against those specifically
  • Training your staff remember, because it is about them
—Contact

Tell us what you are worried about

One call is enough to scope an engagement. We will tell you whether we are the right firm for it, and what it will cost.

[email protected]
  • Target

    What should be tested: an application, a network, an Active Directory estate, a cloud tenant, an AI system, your people, or all of it.

  • Concern

    What you are worried about: ransomware, a regulator, a customer security questionnaire, a breach you already had.

  • Timing

    Any deadline you are working against: an audit date, a release, a board meeting, a contract renewal.

  • Constraints

    Production systems that cannot be touched, change freezes, legacy you already know about, or testing you have had before.