AI Engineer & Full-Stack Automation Developer based in Berlin.
I build reliable AI-assisted engineering systems, agentic workflows, secure APIs, CMS platforms, and automation-heavy B2B products. My work connects LLM orchestration with maintainable production software, clear quality gates, and human-in-the-loop control.
- Agentic systems: orchestration, tool calling, structured outputs, MCP-style integrations, human-in-the-loop validation
- Backend: Python, Django, Django REST Framework, FastAPI, Celery, Redis, PostgreSQL
- Full stack: TypeScript, Next.js, React, Prisma, Sanity CMS
- Delivery: Docker, Kubernetes, GitHub Actions, CI/CD, testing, security
- AI Harness — a policy-governed system for running verified software tasks in local Git repositories, with task isolation, recovery checkpoints, quality gates, and explicit human approvals.
- Clean aus Berlin — a production website case study built with Next.js, TypeScript, privacy-aware integrations, and Cloudflare-ready infrastructure.
- Blockchain Truth Machine — a Web3 case study connecting Django, wallet interactions, blockchain transactions, and IPFS storage.
- Primest Lead Webhook — a Django service for filtering, validating, transforming, and forwarding lead data.
- FHIR Patients — a React and TypeScript application for browsing patient data from a FHIR R4 API.
- Authored 161 merged PRs across two production projects within six months at Primest GmbH, contributing to 98 product and backend issues.
- Built Django REST Framework modules and 10+ endpoints supporting 3,000+ users at Reiner Lemoine Institut.
- Improved API and database performance by 25–40% and maintained 92% backend test coverage.
- Built production workflows with Celery, Redis, webhooks, partner APIs, RBAC, CMS automation, CI/CD, and security controls.