Security review. Find the weak points before someone else does.
We review your application code, access controls, secrets handling and infrastructure, and fix what matters most first.
For teams preparing for a customer security questionnaire, a launch or new rules on data and AI.
Security review
We review authentication and authorization, data exposure between users, secrets, dependencies, infrastructure configuration and logging, including risks specific to AI features such as prompt injection and data leaking through models. You get a ranked findings report and, if you want, the fixes.
- (01)Access-control and data-exposure checks
- (02)Secrets and dependency review
- (03)AI-specific risks covered
- (04)Ranked, actionable findings
Tools we use for this.
- OWASP ASVS
- OWASP Top 10 for LLM applications
- Dependency scanners
- Cloud configuration review
- Static analysis
- Logging review
We choose the stack that fits your requirements. Tool names describe our skills, not partnerships.
Setup.
1–2 weeks.
One clearly bounded piece of work, such as a system configured or a review completed, handed over ready to use.
Included
- (01)Scope and threat model
- (02)Code and configuration review
- (03)Access-control testing between users and roles
- (04)AI-specific risk review
- (05)Findings report ranked by risk
- (06)Fix plan and a retest of fixed items
Not included
- (01)Formal penetration testing or certification (we can bring in an accredited partner)
- (02)Physical and social-engineering tests
- (03)Compliance sign-off
How it runs
- (01) · Days 1–2
Scope
Systems, data, threats and the rules that apply.
- (02) · Week 1
Review
Code, access, infrastructure and AI features.
- (03) · Week 2
Fix and retest
Fix the top risks and confirm they are closed.
Also available as: Build & Support (Staged, ongoing). Compare the formats.
You might also
look at these.
Is this a penetration test?
No. It is a hands-on security review of code and configuration. If you need a formal penetration test or certification, we can coordinate an accredited specialist.
Do you cover AI features?
Yes: prompt injection, data leaking between users, unsafe tool use by agents, and logging of model inputs and outputs.
Will you fix what you find?
If you want us to. The report ranks findings so the fixes can be scoped and priced.