- Market
- Saudi Arabia
- Written for
- CISO, CTO
- Decision
- Measure an AWS estate against a security baseline
- Languages
- EN · AR
Cybersecurity credentials
NCA registration
Registered cybersecurity service provider

ISO/IEC 27001
ISO/IEC 27001 certified
Coverage
The areas of an AWS account a review has to cover
The baseline groups the account into areas, and a weakness in one rarely stays there. These are the areas we read, what we examine in each, and what a gap there means.
Identity and access management
Root account use, multi-factor coverage, over-broad policies, unused credentials and the roles that can escalate their own privileges.
What a gap there means: This is where the account is won or lost; a broad policy is standing permission an attacker inherits on the first foothold.
Network exposure
Security groups open to the internet, public subnets, exposed databases and load balancers, and endpoints reachable that should be private.
What a gap there means: A public database or an open management port is exposure that needs no exploit to be a serious finding.
Logging and monitoring
Whether audit logging is on in every region, protected from deletion, and delivered somewhere it would be seen — not merely enabled.
What a gap there means: Without it, an incident cannot be reconstructed and a quiet intrusion leaves no record to find later.
Encryption and key handling
Data encrypted at rest and in transit, who can use and manage the keys, and whether a key policy grants more than the workload needs.
What a gap there means: A key any role can use is not protecting the data; the control is the policy on the key, not the encryption itself.
Data exposure
Storage buckets, snapshots and images that are public or shared, and data stores reachable across account boundaries.
What a gap there means: A single public bucket is the most common way an estate leaks, and it is a configuration setting, not an attack.
Review or test
What a posture review answers that a penetration test does not
A posture assessment and a cloud penetration test answer adjacent but different questions, and buying one while meaning the other is a common and expensive mistake. The assessment reads the configuration and reports what the account is set to allow, measured against a baseline. It is comprehensive by construction — it sees every bucket, every policy, every open security group — because it reads the state rather than trying to reach it. It does not, and cannot, tell you which of those findings an attacker would actually chain into access.
The sensible order for most organisations is the review first. It is faster, it needs only a read-only role, and it surfaces the whole set of misconfigurations at once — a public database, a role that can escalate, logging switched off in a region nobody uses. Fixing those is cheaper before a test than as test findings, and it lets the later penetration test spend its time on the paths that survive a clean baseline rather than on the open door anyone could have closed from a report.
How it runs
How the account is reviewed, from a read-only role outward
1 · A read-only audit role
You grant a role scoped to read configuration and nothing to change it. The review never needs write access, and saying so plainly is part of how the engagement is trusted inside a regulated organisation. Access that cannot alter the estate it is examining.
2 · Collect state across accounts and regions
The configuration is read across every account in the organisation and every region, because the forgotten resource is almost always in the account or region nobody actively uses. A complete picture, not a sample of the main account.
3 · Measure against the baseline and reason about impact
Each area compared to the well-architected security pillar and the CIS benchmark, then each finding weighed by what it actually exposes rather than by the benchmark's generic severity — a public bucket of logos is not a public bucket of customer records. Findings ranked by exposure, not by a checklist's default score.
4 · Remediation and what still needs a test
Each finding written with the exact change that fixes it, and a plain statement of what a configuration review cannot establish — so the areas that genuinely need a penetration test are named rather than left implied. A fix list and an honest boundary of what the review covered.
Our role
What the review produces and what your cloud team then owns
- 01
We read, we do not change
The assessment runs from a read-only role and never alters the estate. What it produces is a picture measured against a baseline you agreed, not a set of changes made on your behalf.
- 02
We rank by exposure
Findings are ordered by what the misconfiguration actually exposes in your account, so the team fixes the public data store before the benchmark note that carries no real risk in your context.
- 03
Your cloud team applies the fixes
The remediation is written to be applied by the people who run the account, in your infrastructure-as-code where you use it, so the fix is durable rather than a manual change that drifts back at the next deployment.
Which findings to remediate, and how to fold the fixes into your infrastructure-as-code, are your team's to run; we establish the posture against a named baseline, rank it by real exposure, and mark the boundary where a penetration test has to take over.
Common questions
What buyers ask
Is an AWS security assessment the same as a cloud penetration test?+
No, and the difference decides which one your situation needs. An assessment reads the configuration from a read-only role and reports, comprehensively, what the account is set to allow measured against a baseline — every bucket, every policy, every open group. A penetration test attempts to reach production data by chaining what it finds, and it establishes what an attacker could actually do rather than what the settings permit. The assessment sees everything but proves nothing is reachable; the test proves a path but only sees what it walks. For most organisations the review comes first, because it is cheaper to fix an open door from a report than to have it found as a test finding.
Do you need administrative access to our AWS account?+
No — a read-only audit role is the whole of what the assessment needs, and it is deliberate that it cannot change anything. The role is scoped to read configuration across your accounts and regions and nothing to modify the estate, which is both safer for you and cleaner for the record inside a regulated organisation, where an external party holding write access to production is a finding in itself. If a specific check needs data a read-only role cannot see, we name it and you decide whether to provide it, rather than asking for broad access up front.
How is this different from the security checks AWS gives us for free?+
The provider's own tools are a genuine input and we read their output, but they report findings without a judgement of what each one means for you. They will flag a public bucket identically whether it holds marketing images or customer records, note a broad policy without tracing where it actually reaches, and produce a long list that is comprehensive and unprioritised. The assessment adds the reasoning the tools cannot: which findings chain together, which matter given what your account actually holds, and which are noise in your context. The value is not seeing the findings — it is knowing which five of the four hundred to fix this week.
In this cluster
Next step
Grant a read-only audit role
A read-only assessment of the AWS estate against a named baseline, returning a prioritised posture report with the remediation for each finding and the areas that need a penetration test instead.
Grant a read-only audit role
What happens next
- 1We reply within one business day.
- 2A 30-minute scoping call with a senior consultant.
- 3A written proposal with scope, team, and timeline.
Or reach us directly
