FINTECH & PAYMENTS
Client confidential

Crypto-Friendly Neobank

FINTECH CRYPTO-FRIENDLY NEOBANK GLOBAL

A full AWS infrastructure review across compute, storage, database, networking, and IAM. The kind of audit that finds the quiet defaults nobody revisits after launch: a public bucket, an unencrypted volume, a root account still in daily use.

Engagement type
AWS Cloud Security Audit. Full infrastructure review.
Scope
Complete AWS cloud environment including compute, storage, database, networking, and IAM.
/ context

Cloud environments do not get misconfigured. They accumulate.

Almost nobody sets out to leave a storage bucket open to the internet. What happens instead is that a bucket is made public for a legitimate reason during a build, the reason goes away, and the setting does not. Multiply that across a few years of shipping quickly and the environment carries a set of decisions nobody currently remembers making.

AWS operates a shared responsibility model. Amazon secures the infrastructure the cloud runs on. Everything configured inside it, who has access, what is encrypted, what is reachable from the internet, belongs to the customer. Most cloud incidents are not breaches of the platform. They are customer configurations behaving exactly as configured.

For a neobank the exposure is not abstract. The data sitting in those buckets and volumes is user financial data, and the identity layer governing access to it is the same layer that governs access to production. A full review of the environment is the only way to see the accumulated position rather than the most recent change.

/ findings overview

Forty-two findings across the full AWS estate.

42
Total findings

Across compute, storage, database, networking, and IAM.

1
Critical severity

Cloud storage publicly readable on the open internet.

8
Critical and high

The set prioritized for immediate remediation.

Severity distribution

1 Critical7 High7 Medium24 Low2 Informational

Outcome

Critical public data exposure was resolved. IAM, encryption, and network security posture improved significantly on retest.

/ approach

How the audit ran

The review covered the complete AWS environment: compute, storage, database, networking, and identity and access management. Nothing was scoped out, which matters for an audit of this type because the findings that combine into real risk usually sit in different services.

Configuration review of a cloud estate is a different discipline from application penetration testing. The work is enumerating what exists, comparing each resource against a hardening baseline, and then determining which deviations are genuinely exploitable in this particular environment rather than theoretically undesirable.

Findings were rated by severity and re-examined after remediation. The retest is what distinguishes the reported outcome from a list of recommendations: the critical public data exposure was confirmed closed rather than assumed closed.

/ what we found

Key risk areas identified

Seven categories of exposure across the cloud environment, from open data to unmanaged identity.

01

Publicly readable storage buckets

User and operational data exposed to the open internet with no access restriction.

A publicly readable bucket needs no attack. Cloud storage is enumerable, indexes get crawled, and researchers and attackers both scan continuously for open buckets. If it holds user data, the exposure begins the moment the setting is applied rather than the moment somebody exploits it.

02

Unencrypted volumes and snapshots

EBS volumes and database snapshots holding sensitive user financial data, stored in the clear.

Encryption at rest is what protects data from access paths that bypass the application entirely: a copied snapshot, a detached volume, a backup restored somewhere it should not be. Snapshots are the commonly missed half, because they inherit their encryption state at creation and then persist long after the volume is gone.

03

Root account used without MFA

The highest-privilege credentials in the account accessed and used without a second factor.

The AWS root account can do things no other identity can, including actions that cannot be restricted by policy. Standard practice is to secure it with hardware or app-based MFA and then stop using it for daily work entirely. Active use without MFA inverts that.

04

User accounts without MFA

Multi-factor authentication missing across user identities in the AWS environment.

MFA gaps across ordinary users matter because credentials leak constantly, through reuse, phishing, and third-party breaches. A second factor is what stops a leaked password from becoming account access.

05

Clear-text traffic at the load balancer

User traffic transmitted over HTTP and open to interception in transit.

Traffic served over plain HTTP is readable by anything positioned between the user and the load balancer. For a financial service that includes session tokens and account data moving in the clear, on a path the user cannot inspect.

06

Publicly accessible database clusters

Clusters reachable from the internet, unencrypted, with backup retention too short for reliable recovery.

A database cluster reachable from the internet has its entire attack surface exposed to everyone rather than to the application. Combined with encryption being off and backup retention being short, the result is a system that is easier to reach, easier to read, and harder to recover.

07

IAM and network hygiene

Policy misconfigurations, unused credentials, weak password policies, and unused security groups.

Identity hygiene is what determines blast radius. Over-permissive policies decide how far an attacker gets after a single account falls, unused credentials are access nobody is watching, and unused security groups are open paths nobody remembers opening.

/ takeaways

What this audit says about fintech on AWS

01

The critical finding was a setting, not a vulnerability

Publicly readable storage required no exploit and no skill to reach. Anyone who found the bucket could read it. The most severe issue in a cloud environment is frequently the most boring one.

02

Twenty-four low severity findings are the actual story

Individually each was minor. Collectively they describe an environment where hardening was not a routine step, which is the condition that produces the critical finding in the first place.

03

MFA on the root account is not a checkbox

Root holds privileges no IAM policy can restrict. Using it without a second factor means the strongest credentials in the account are protected by a password alone.

04

Encryption has to cover both states

Unencrypted volumes and snapshots address data at rest. A load balancer serving clear-text HTTP addresses data in transit. Getting one right and not the other leaves the data exposed, just at a different point.

/ relevant for

If this looks like your environment, it probably behaves like it too.

Fintech companies on AWS or any major cloud.

Neobanks and digital financial services with cloud-first architecture.

Companies pre-fundraising or pre-audit needing a comprehensive cloud hygiene review.

/ common questions

Questions this engagement usually raises.

What is a cloud security audit?
A structured review of a cloud environment's configuration against a hardening baseline, covering identity, storage, compute, networking, and databases. It identifies exposure created by how resources are configured rather than by flaws in application code.
What is the AWS shared responsibility model?
AWS secures the underlying infrastructure. The customer secures everything they configure within it, including access control, encryption settings, and network exposure. Most cloud data incidents fall on the customer side of that line.
Is a cloud audit the same as a penetration test?
No. A penetration test attacks a running application to find exploitable flaws. A cloud audit examines how the environment is configured to find exposure that requires no exploitation at all. Environments generally need both.
Why do unencrypted snapshots matter if the volume is secure?
Snapshots are independent copies that can be shared, restored, or retained long after the original volume is deleted. An unencrypted snapshot is a readable copy of the data sitting outside the controls protecting the live system.
What made the critical finding critical?
Cloud storage buckets were publicly readable, exposing user and operational data to the open internet. No credentials or exploitation were required to reach it.
How often should a cloud environment be audited?
Common practice layers three cadences rather than settling on one number. Continuous automated posture monitoring runs constantly and catches configuration drift as it happens. A comprehensive manual review runs at least annually, which is the floor most frameworks assume: PCI DSS calls for testing at least every twelve months and after any significant change, while SOC 2 and ISO 27001 both operate on annual cycles. Regulated and fast-moving environments usually move to quarterly. Separately from the calendar, certain events should trigger a review regardless of when the last one ran: a major architecture change, a new account or region, entry into a new compliance scope, an acquisition, or an incident. Cloud configuration drifts continuously, so a point-in-time review begins aging the day it is delivered.
/ more engagements

Other work.

Start here

Think like them.
Before they do

Book a 30-minute walkthrough. We'll point CredShields at a scoped asset and show you live findings by the end of the call.

Secure your protocol today

Don't wait for a
security incident.

Get your comprehensive security audit from the team trusted by 200+ protocols and enterprises worldwide. Fast turnaround. Proven track record. Direct access to senior security engineers.

Fixed-Fee Pricing
No engineer-hour billing
Audit-Ready by Default
SOC 2, ISO, PCI, HIPAA
Engineer-Validated
Not scanner output