Junior Infrastructure Engineer (full-time · long-term · remote)

Hi, we are DeepMetis 🚀 a developer-founded, Berlin-based, 100% remote AI company. We're 26 people (10 devs in your future team, 8 in other teams) operating across Germany and Tunisia, and we build cross-sector solutions with a strong focus on AI, Machine Learning, and custom software engineering.

Our culture matters as much as our code. We're supportive, curious and open-minded. We joke around a lot, but we're serious about the quality of our work. We love cats, and our Slack sometimes looks like a meme museum.

We run the production systems our AI products depend on. We're looking for someone early in their career who wants to go deep on infrastructure — not babysit dashboards, but understand why things work and why they break

What your day-to-day work looks like:

• 🧑‍💻 Keep production healthy: Help operate and improve our Kubernetes clusters — deployments, scaling, the occasional "why is this pod unhappy" investigation. You won't be alone; you'll pair with senior engineers until you're confident flying solo.
• 🔍 Diagnose, don't just react: When something's down, the obvious cause is rarely the real one. You'll learn to trace symptoms back to root causes — "the API is timing out, but actually it's a DNS issue three layers down." We don't expect you to have this instinct yet; we expect you to want it.
• 🚀 Own the delivery pipeline: Work on our CI/CD (GitLab) and GitOps (ArgoCD) workflows — making deployments faster, safer, and more boring (boring is good in infra).
• 🌐 Get your hands dirty with networking & nginx: Routing, ingress, TLS, load balancing. The unglamorous plumbing that everything else relies on.
• 🤝 Grow with your team: Join design discussions, ask "why are we doing it this way," and build your judgment by being in the room when the hard calls get made.

You'll probably love this role if you bring most of the following:

• 🗣️ Fluent English: we're an English-only company.
• 🧱 Solid foundations: you understand how computers, networks, and the Linux command line actually work — not just which buttons to press. We'd rather have strong fundamentals than a long tool list.
• 💼 0–2 years of experience: internships and serious side projects count (roughly half-weight). If you've kept a homelab, broken your own k8s cluster, and fixed it — that's a real signal.
• ☸️ Some Kubernetes exposure: you don't need to be an expert, but kubectl shouldn't be scary. You've at least deployed something to a cluster.
• 🧰 Familiarity with CI/CD and Git: you understand what a pipeline does and why GitOps is more than a buzzword.
• 🧠 A diagnostic mindset in the making: you're the kind of person who can't leave a bug alone, who asks "but why," and who isn't satisfied with "I restarted it and it works now."
• 😌 Calm under pressure: when production is on fire, you slow down instead of panicking. (To be clear: business hours only, no night shifts — see below.)
• 💬 Good communicator: comfortable in async text (Slack) and on calls. You don't disappear for a week into a branch.
• 😸 Bonus: you have a cat. (Optional. We don't discriminate against dog people. Much.)

Honest expectations:

This is production infrastructure, so there's real responsibility — but we've set it up to be humane. Business hours only. No night shifts, no 3 a.m. pages. When things break during the day, you'll learn the most by being there for it, with senior engineers alongside you. "Under pressure" here means staying composed during a daytime incident, not living on a pager and most importantly not being afraid.

Top working conditions! Enjoy an amazing working space with great benefits:

• 🌴 Remote work: Work from home or wherever you like; we are a 100% remote company.
• 🕚 Flexible working hours: Within business hours, work when you feel sharpest.
• ⭐ Challenging projects: Real systems, real impact, real things to learn.
• 😄 Amazing, creative team: Supportive, curious, open-minded. We joke around a lot but we're serious about the quality of our work.
• 💡 Continuous learning budget: We pay for your experiments and your side quests.
• 🌱 Fault-tolerant culture: Mistakes are part of learning. We make them, we learn, we move on. No blaming, no negative feelings — maybe a few jokes.
• ...and yes, you do get a Claude Code subscription.

We don't need a 5-page CV. We care about how you think and what you've actually built.

If this sounds like your kind of place, hit apply.

We're looking forward to meeting you!

— The DeepMetis team