I build digital products end-to-end — from technical discovery to production and scale. I focus on the parts that usually cost teams the most: authentication done right, payments and document workflows that don't break on Fridays, mobile apps that share a core with the web, and architecture that survives the team that comes after me.

Lately I've been going deep into process automation with self-hosted AI: real operational problems solved with n8n, Docker and applied AI on infrastructure I design, deploy and maintain myself.

Available for freelance contracts and high-impact technical collaborations. Remote, EU timezone preferred.

Happy to share a technical breakdown of any of these (architecture, key decisions, trade-offs resolved).

I run my own automation hub — n8n + Docker + AI, self-hosted on a Mac Mini behind a Cloudflare Tunnel, with zero third-party services for sensitive data. On top of that infrastructure, two systems running in production today:

1 · AI email assistant

• Classifies every incoming email with AI (important / personal / lead / spam) and tags it autonomously, regardless of language.
• Drafts reply suggestions for leads, ready for review and send.
• End-to-end invoice handling: detects incoming and outgoing invoices, reads the PDF, verifies recipient/NIF, and prepares the accountant draft with the attachment.

2 · Performance simulation platform · Amentibus Cultui (WoW guild)

• Moved heavy compute (SimulationCraft runs + external APIs) from cloud to my own hardware, orchestrated with n8n — direct server cost reduction, web stays as a thin read layer.
• Nightly pipeline computes performance for the whole roster and persists results; on-demand button on the web returns a fresh analysis instantly.
• Droptimizer-style feature: simulates which item upgrades each character the most, comparing dozens of combinations in a single pass.
• Built with engineering rigor: concurrency control and queues, secrets in credentials (never in code), Git versioning and automated backups.

The differential: I take real problems — daily paperwork, runaway infrastructure costs — and solve them with automation + applied AI on infrastructure I own. Not demos. Systems that run themselves, every day.

Frontend       Next.js 14 (App Router · RSC · Server Actions) · React · TypeScript
               TanStack Query/Table · Tailwind · React Hook Form + Zod
Mobile         React Native · Expo · iOS + Android in production
Backend        Node.js · REST/GraphQL · Webhooks · background jobs
Data           Prisma · PostgreSQL · Supabase (Auth · DB · Storage · Realtime · Edge)
Payments       Stripe (Checkout · subscriptions · billing portal · webhooks)
Documents      DocuSeal (e-signature · document workflows)
Auth           Magic Link · OAuth · JWT · RBAC · OWASP-aligned hardening
Web3           EVM ecosystem · wallet auth · on-chain read/write · off-chain sync
Automation     n8n (self-hosted) · Docker · Cloudflare Tunnel · AI workflows
Architecture   Hexagonal · Clean Code · SOLID · tactical DDD
Quality        Unit + integration + E2E testing · GitHub Actions CI/CD · Vercel

• End-to-end product development: discovery → MVP → production → scale.
• High-impact features: auth flows, payments and subscriptions, document signing, admin panels, mobile apps sharing the same backend core.
• AI-powered automation of operational processes — replacing recurring manual work with reliable, observable systems.
• Refactors of legacy code with measurable improvements in performance and maintainability.
• Comfortable working with product, design and business — not just other engineers.