I am Nawinn, born 1984, based near Frankfurt. I took my first PC apart as a child, started coding at 16, and worked as an application developer and server administrator from roughly 2002 to 2011 — at CPC-IT, SDS Informationstechnik and a continuing-education center. Stack then: C#, C++, interface programming, Windows servers, Linux backend configuration.
Then came a professional pause from IT — other jobs, continuing strong interest in technology, and a close watch on what has been happening: Hetzner as a serious AWS alternative, local LLMs like Llama 3.3, Astro instead of Next.js-overkill, Tailwind v4, Cal.com instead of Calendly. I watched — I did not yet ship.
Since 2025 I am a full-time developer again — this time with Claude Code as a constant co-pilot. This is not a gimmick. It is a fundamentally different working mode. Architectures that once took a two-person team a month, I now build in two weeks. Not because the AI "can do it all", but because it frees me from boilerplate so I have time to build the right thing.
Why "workshop"? Because software is a craft. Every project has its particularities. No industrial "sprint cycles" for one-person engagements — instead carefully-tuned work steps, clean documentation, traceable decisions.
Why "stackschmiede"? Because I cover the full stack — from server setup to UI. Backend, frontend, infrastructure, increasingly AI integration with local models. A small team in the background for emergency backup.
Plainly said: I am at the start of my freelance career. For the first 3-5 projects I therefore offer pilot terms: fair entry pricing in exchange for permission to use the project as a public reference (with an NDA on internals when needed). After that I return to regular rates. A deliberate win-win — you pay less, I build my reference list.
I write code the next person — or me in six months — understands without puzzles. Named variables, documented invariants, no magic. Exactly what AI assistants were not designed for: producing clean, readable context.
Claude Code is my co-pilot — not my autopilot. 8-9 years as a C#/C++ application developer and server admin stand behind every architecture decision. The AI writes faster. The responsibility stays mine.
OpenAI, Anthropic and Google are a competitive advantage as internal dev tools — Claude Code as co-pilot speeds my development substantially. For your shipped product I deliberately choose on-prem LLMs (Llama, Mistral) on your Hetzner server. That is your business-continuity insurance, not mine: pricing, APIs and regions stay under your control.
I host in DE/EU, use open formats, build on widely-adopted standards. No Vercel dependency, no Firebase lock-in. Sometimes harder initially — pays off the first time a vendor conflict arises.
Three channels, one contact. Reply within 24 hours on business days.