Senior software engineer with years of backend and full-stack work, on information security, e-commerce, education and direct-selling platforms. I turn complex business rules into clean, tested, maintainable systems. Today I also use AI in a disciplined, engineered way that makes that work faster without making it sloppier.
I've worked across industries, but the job is the same everywhere: understand the domain, model it well, and ship software people can rely on and keep building on.
Laravel and Node.js services, REST APIs, queues and relational data design. Complex rules such as commissions, orders, compliance and accounting are modelled explicitly and tested where it matters.
Domain-Driven Design, Clean Architecture, SOLID and design patterns. Multi-tenant SaaS, refactoring legacy code into clear layers, and architecture decisions written down.
React, Next.js, Inertia and Livewire frontends and progressive web apps. I own a feature end to end instead of handing over half of it.
Pest and PHPUnit suites, Playwright end-to-end tests, static analysis, Docker, CI/CD pipelines and AWS, including serverless deployments on Lambda.
Direct selling and MLM, information security management, e-commerce and education. I talk to stakeholders in their own terms and turn vague requirements into precise behaviour.
I design the guidelines, skills and verification loops that let coding agents do reliable work. More on that below.
German software house building custom MLM and direct-sales platforms for partner networks worldwide.
Canadian cybersecurity company in Winnipeg. I worked on the platform behind iSecureData Copilot, a SaaS that helps organisations implement and run security frameworks such as ISO 27001 and HIPAA.
Online marketplace that connects students with vetted private tutors, for online and in-person lessons, serving students at home and abroad.
Projects for small businesses and startups: e-commerce stores, progressive web apps and SaaS products.
An agent is only as good as the environment it works in. My engineering background shapes how I use AI: the same standards, enforced by written rules and checks instead of hope. The result is better quality, more speed and fewer bugs, on any kind of software, not only the web.
Current state, decisions and history live in the repo, so every session starts informed.
Reuse is the default path, because duplication is the main failure in AI-written code.
Library behaviour is checked against version-matched docs, and claims against tests and HEAD.
Architecture decisions are recorded. A stale document counts as a bug.
Read-only guards on real data, approval before remote writes, architecture tests.
Every mistake caught once is written down, so no agent repeats it.
Settle the design first. The outcome is written back into docs and skills.
Read the codebase carefully before planning anything.
A master plan split into small steps with acceptance criteria.
Step by step, tested at every step, with progress kept on disk.
Decisions, docs and current state updated, so the next session starts warm.
I'm happy to talk about engineering roles, projects, or bringing AI into a team's development the right way.