Erhalte mehr Antworten von Arbeitgebern
Versende in nur wenigen Minuten einen passgenauen Lebenslauf.
Ocumeda sucht eine/n Staff Software Engineer (m/f/d) in München. Sie gestalten die Full-Stack-Plattform (Frontend, Backend, Daten, AWS), definieren Testing, Guardrails und eine agile Coding-Kultur mit Fokus auf Qualität und Sicherheit.
Sie arbeiten eigenständig an Architekturentscheidungen, treiben Produkt-Verbesserungen voran und sorgen dafür, dass Lösungen operabel und auditierbar bleiben. Eine anspruchsvolle, international ausgerichtete Schnittstelle erwartet Sie.
As Staff Software Engineer (m/f/d) you are the engineer who makes hard, unclear or broken things tractable. You know how our platform works end to end and you shape it as you go. Your focus is our full-stack platform: the web application our doctors, our optician partners, and their customers use daily, and the services behind it. This is an individual contributor role. No direct reports, real technical authority: you set direction on architecture and engineering practice, and you build in the same codebase as everyone else in the team. Staff/Senior here means deeper, not further from the code. Your job is to keep our systems fast to iterate, reliable in production, and ready for international expansion and scaling. Own the architecture of our full-stack platform - frontend, backend, data, AWS - and evolve it under real product pressure instead of rewriting it. Own our testing strategy: where we invest in unit, integration and end-to-end coverage, what we deliberately do not test, and how tests stay fast enough that nobody skips them. Push our agentic coding practice forward: tooling, guardrails and testing discipline that make us faster without putting unverified code into a live product. Shape scope together with product: surface the cheap 80% solution, and say clearly when something is not worth building. Build so that what ships is operable and auditable by design (e.g. ISO 27001, ISO 13485), without turning engineering into paperwork. Go wherever the problem is - a scaling bottleneck one week, a broken integration the next - you fix it properly.