Available for new projects

I fix and secure AI apps before they break in production.

Teams ship AI features and whole apps that work in the demo and fall over on the hard parts — broken auth, leaking secrets, prompt-injection holes, half-wired integrations. I do the other half: make them work, and make them safe.

Production down or a live security issue? Get urgent help now

Same-day and next-day emergency slots available (priced for the drop-everything turnaround). Send the email and I'll come back fast.

auth.py found: plaintext auth · critical fixed · constant-time
# AI-generated login — works in the demo, unsafe in prod if user.password == request.form["pw"]: if bcrypt.checkpw(request.form["pw"].encode(), user.pw_hash): return issue_session(user)
What I do

The other half of building with AI.

AI tools get you to a demo fast. I take it the rest of the way — reliable, secure, and tested.

rescue

AI-generated code rescue

Debug and harden the auth, database, API, validation, deploy, and test gaps that Lovable / Replit / Cursor leave behind.

security

LLM & app security

Prompt-injection probes, OWASP-LLM-2025-aligned reviews, secrets/misconfig checks, and secure-code review.

integrate

AI integrations & automation

LLM extraction/classification wired into APIs, ERP/CRM systems, and n8n/Make workflows — with retries, idempotency, and validation.

build

Custom AI tools

RAG assistants, document extraction, support triage, lead-gen pipelines, and the internal automation around them.

Proof you can run

Tested reference projects — not screenshots.

Each is a small, honest capability demo with a clear threat model, real tests, and a measurable before/after you can reproduce in one command.

Fix & secure AI-generated apps

24 tests · AST fixer

Detects real issues (debug-in-prod, missing auth, plaintext passwords, hardcoded secrets, injection-prone DB calls) via AST analysis, patches supported Python, and rebuilds auth the safe way — with a before/after diff.

$git clone https://github.com/clira-dev/vibe-code-rescue && cd vibe-code-rescue && pip install -r requirements.txt && pytest
View on GitHub

LLM / GenAI security audit

OWASP-LLM 2025 · 22 tests

A black-box red-team aligned to the OWASP LLM Top-10 (2025): prompt injection, sensitive-info disclosure, output handling, excessive agency, system-prompt leakage. Surfaces 11 findings on a vulnerable target, 0 on the hardened one.

$git clone https://github.com/clira-dev/llm-security-audit && cd llm-security-audit && pip install -r requirements.txt && pytest
View on GitHub

AI document extraction → validated JSON

8-doc gold set · 27 tests

Turns messy invoices/receipts into clean, validated JSON with arithmetic reconciliation and confidence scoring — low-confidence docs route to human review. Quality is measured on a labelled gold set, not claimed.

$git clone https://github.com/clira-dev/ai-doc-extractor && cd ai-doc-extractor && pip install -r requirements.txt && python3 evaluator.py
View on GitHub

AI ↔ ERP/CRM integration

idempotent · 29 tests

A production-shaped pipeline: incoming order/email → AI-structured → validated → synced through a resilient client with content-addressed idempotency (corrected re-sends never double-post), retries with backoff, and typed errors.

$git clone https://github.com/clira-dev/ai-erp-crm-integration && cd ai-erp-crm-integration && pip install -r requirements.txt && pytest
View on GitHub

Like what you see? Start a project

How I work

Scoped, tested, no surprises.

01 / scope

We agree what "done" means

A small, clear threat model or spec up front — checkable acceptance criteria, not vibes.

02 / build

I build it, with tests

Real, documented work and a measurable before/after — never a black box.

03 / verify

You see it working

Verified against your case before sign-off, with one revision round included.

Start a project

Send one repo or AI workflow and what "done" looks like. I'll reply with a focused scope, the highest-risk issues, and a price.

Your code stays confidential — NDA on request. I never store or redistribute client code.

Production down or a live security issue?

Get urgent help now

Same-day and next-day emergency slots available (priced for the drop-everything turnaround). Send the email and I'll come back fast.

Free, no-obligation scope within one business day.