$ ~/secure-code-audit

// found → fixed → verified

Your codebase, audited and fixed.

I'm a PhD in AI and cybersecurity. I audit web app codebases against the OWASP Top 10, pen-test your live endpoints, then fix what I find and hand back working code. Fixed scope. Fixed price. No retainer required.

[ request an audit → ]

scoped per project · fixed price, no hourly · 50% deposit

01 // the problem

You don't know if your app is secure — and finding out is expensive.

Pentest firms quote five figures and hand you a PDF. Automated scanners bury you in false positives and miss the bugs that actually matter. Meanwhile a client security questionnaire or a compliance deadline is sitting in your inbox with a date on it.

report, not a fix

a pentest ends where your work begins.

scanner noise

false positives, and it still misses access control.

deadline landed

a questionnaire or audit just became your problem.

built fast

security was a "later" problem. later is now.

02 // coverage

What gets checked.

Security isn't one category. I audit against the three properties that define it — confidentiality, integrity, availability — because an app can be locked down and still be broken.

confidentiality

can anyone see data they shouldn't?

  • injection: SQL, NoSQL, command, template
  • broken access control: IDOR, missing role checks, horizontal escalation
  • auth and session handling: weak tokens, fixation, insecure resets
  • exposed secrets: keys and credentials in code and git history
integrity

can anyone change data they shouldn't?

  • business logic flaws: price and quantity manipulation, workflow bypass
  • race conditions: double-spend, duplicate submission
  • input validation: unsanitized data reaching storage
  • supply chain: unpinned dependencies, unverified lockfiles
availability

can the app be degraded or taken down?

  • slow queries and unbounded operations under real load
  • missing rate limits on expensive endpoints
  • dependencies with known DoS CVEs
  • error handling that cascades instead of failing gracefully

Most audits stop at the first one.

03 // deliverables

Every finding, and the code that closes it.

findings report
every issue, severity-rated, file and line, mapped to OWASP/CWE
fixed code
patches delivered as a reviewable branch, grouped by severity
verification
every fix retested and confirmed closed
hardening guide
the checklist and config changes that keep it closed

No vague recommendations. Working code, with the holes closed.

04 // process

How an engagement runs.

01

Scope & access. Kickoff call: we agree exactly which repos, services and endpoints are in scope, and I quote a fixed price and timeline. We both sign it — that document is the contract.

02

Audit. Manual code review plus tooling across the codebase — injection, access control, auth and session handling, secrets, dependencies, misconfiguration — alongside live pen testing of your endpoints.

03

Report. Prioritized findings: what's wrong, where, why it matters, what it maps to. Critical issues reach you the day I find them, not at delivery.

04

Remediation & verification. I fix the confirmed issues and deliver them for review, grouped by severity so critical fixes merge first. Each one is retested and confirmed closed, then a walkthrough call to hand over.

05 // whoami

$ whoami

Dr. Kamal Bella, founder

Dr. Kamal Bella. PhD in AI and cybersecurity. Based in Morocco. I build products, not just papers. What that means for your audit:

security

PhD-level background in cybersecurity. I find the flaws tools don't flag.

real fixes

I don't hand you a list. I write the patch and prove it closed the hole.

root cause

fixed at the source, so the same class of bug doesn't come back.

upwork earnings
$50,000+
rating
5.0 / 5.0
total hours
1,000+

"Kamal was quick, got it 100% right away and responded immediately to requests. I would tell you he was perfect based on my interactions with him and I am massive pain in the ass."

— Mark

Not a scanner. Someone who reads your code.

06 // pricing

Fixed price. No hourly.

Scope and price are agreed before anything starts. No creep, no meter.

scope
priced per project after a free scoping call
basis
codebase size, stack, and endpoint count
timeline
set at scoping, fixed
deposit
50% — reserves your audit slot
balance
50% — due before handover; fixes released on payment

07 // faq

Q: What access do you need?

A: Read access to the codebase, and permission to test the endpoints in scope — staging where possible, production only with written sign-off. Fixes come back as a reviewable branch or patch set that you merge yourself. I never need write access to your main branch.

Q: What languages and frameworks do you cover?

A: Any stack. The vulnerability classes are the same everywhere — injection, broken access control, auth handling, secrets, misconfiguration — and reading unfamiliar code carefully is the job.

Q: Is this a penetration test?

A: It's both halves. I read the source code, which finds logic and access-control flaws a black-box test can't see, and I test your live endpoints, which finds what only shows up at runtime. Most services do one or the other.

Q: Will this make me SOC 2 / PCI / HIPAA compliant?

A: I audit against whichever framework you're working toward and fix the technical gaps it requires. I'm not a certifying auditor, so I can't issue an attestation — but you go into that audit with the code-level findings already closed.

Q: What if you find something critical mid-audit?

A: You hear about it that day, before the report. Critical issues don't wait for a deliverable.

Q: What happens after the fixes ship?

A: You own the fixed code and the report outright. Re-audits after a major release, and ongoing review, are available separately. No obligation.

08 // apply

Check if it's a fit.

I take a limited number of audits and only ones I'm confident I can close out. First step is a short application. If it's a fit, you get a link to book a scoping call.

// apply --new