Full-Stack Software Engineer with 4+ years of experience building production-grade systems in TypeScript, Node.js, NestJS, and React — plus hands-on JVM (Java/Spring Boot) and a growing focus on shipping autonomous, LLM-powered products end-to-end.
Contact me
Full-Stack Software Engineer with 4+ years of commercial experience building production-grade systems in TypeScript, Node.js, NestJS, and React, alongside hands-on JVM experience through Java/Spring Boot. I own features end-to-end — from REST API design and DDD-based microservices to React/React Native UIs — and increasingly specialize in shipping autonomous, LLM-powered products to production.
From commercial full-stack development to building and running my own autonomous AI product 24/7 in production — here's where I've worked and what I've built.
Designed, built, and operate a fully autonomous, multi-tenant content pipeline running 24/7 in production for multiple branded social accounts. Own the full stack — ingestion, LLM-driven generation (Gemini + local Ollama fallback), automated voiced video rendering (Puppeteer/ffmpeg), and publishing across Instagram, TikTok, YouTube, Threads, and Telegram, including per-account API quota management.
Design, build, and own backend services and REST APIs end-to-end in a microservice architecture (NestJS, TypeScript, PostgreSQL), applying DDD principles — bounded contexts, aggregates, repository patterns. Extend the team's Azure Pipelines for CI/CD, write Jest tests, conduct code reviews, and mentor junior developers.
Delivered JS/TS solutions for complex frontend problems on high-traffic client websites; optimized for performance and cross-browser compatibility.
A mix of production AI systems I've designed and shipped end-to-end, and full-stack academic projects from my MSc at TU Munich.
A fully autonomous, multi-tenant content pipeline running 24/7 in production for multiple branded social accounts. LLM-powered triage, summarization, and image captioning (Gemini, with a local Ollama fallback for quota resilience), automated voiced video rendering, and multi-platform publishing across Instagram, TikTok, YouTube, Threads, and Telegram — built and operated solo, end-to-end.
AI-powered mortgage platform with a scalable Spring Boot backend and LLM integration via RAG architecture — a JVM-based service handling real-time user guidance and complex financial domain logic. TUM academic project.
Multimodal AI requirement-analysis tool combining a React UI with a Python/FastAPI backend — achieved 88% accuracy in automated requirement validation. TUM academic project.
Offline-first German A1 vocabulary app, live on the App Store: spaced-repetition, quiz/game modes, an XP/streak system, and TR/EN/DE localization.
View on App StoreDo you have an idea? Let's discuss it and see what we can do together.
Portfolio.
