AI/ML Engineer (f/m/d)

Your impact is directly measurable: in coding accuracy, in token costs per case and in hospitals that trust every decision because they can trace it.

This is a hands-on engineering role building AI systems in production with real users. It’s not academic research, and it’s not prompt tinkering on the side of a product others build.

How the role sits in the team

You report directly to Christian, our CTO, and work day to day with our engineering team on the agent pipeline, infrastructure and evaluation. With Malte and our product team, you translate what is technically possible into product decisions. Our doctors and coding experts are your sparring partners for clinical quality, and our forward deployed team brings back how the agents perform in real hospitals.

Du berichtest direkt an Christian, unseren CTO, und arbeitest täglich mit unserem Engineering-Team an der Agenten-Pipeline, der Infrastruktur und der Evaluation. Mit Malte und unserem Produktteam übersetzt Du, was technisch möglich ist, in Produktentscheidungen. Unsere Ärzt:innen und Kodierexpert:innen sind Deine Sparringspartner für klinische Qualität, und unser Forward-Deployed-Team bringt zurück, wie die Agenten in echten Kliniken performen.

What you will do

• Build state-of-the-art agents: You build high-performing agentic systems that define the state of the art in healthcare text processing, while staying efficient in token usage and leveraging small models wherever possible.
• Make every decision traceable: You ensure every decision and output of our agents is traced, and you monitor their performance in production.
• Measure what matters: You build the evaluation and benchmarking setup that tells us, on every change, whether our agents got better or worse.
• Bring research to production: You follow new AI/ML approaches, benchmark promising ideas and make sure the best results reach production within weeks.
• Shape the product: You use your AI/ML expertise to evaluate future projects and inform product decisions.

What we are looking for:

• Production experience: At least two years building and shipping ML/AI systems in production.
• LLM and agent expertise: Strong experience with LLMs and agentic frameworks.
• Engineering depth: Strong proficiency in your primary programming language.
• Evaluation mindset: Experience with observability, evaluation and benchmarking of ML systems.
• Product mindset: You care about outcomes that lead to great user experiences and are passionate about using LLMs in real-world scenarios rather than academic research.
• Nice to have: Experience in healthcare, regulated environments or with sensitive data.
• Languages: You speak fluent English. German is a big plus, since the clinical documents our agents process are in German.
• Location: You work from our office in Munich or Berlin.

Missing one or two points, but convinced you’re the right person for this role? Apply anyway. We’re looking forward to meeting you!

What you get

• Work that means something, every day. Your agents are the core of our product and directly help improve the German healthcare system.
• Ownership from week one. You take over real projects with a clear outcome and your name on it, not a list of tasks someone else defined.
• Direct access to the founders. You work day to day with our founders Sophia, a physician and ex-McKinsey consultant, and Christian, who has founded two health tech startups and spent a decade building clinical systems. No layers, no waiting weeks for approvals.
• Trust and freedom. We trust you to solve problems your way, and the team has your back 100%.
• Learning at startup speed. You will see more in one year here than in three years elsewhere, across models, infrastructure, product and customers, with a team that has a proven track record of deploying AI in healthcare.
• One team, in the heart of the city. We work together on-site at WeWork Odeonsplatz in Munich and WeWork in central Berlin. Fully serviced offices with coffee, meeting rooms and phone booths mean zero admin, and the city-center location means short commutes, easy access to customers and investors, and lunch spots around every corner.
• Team events that bring Munich and Berlin together, because the people you build with should also be the people you celebrate with.

What success looks like

• After 3 months, you know our agent pipeline inside out, every agent decision is traced, and you have shipped your first improvement to production.
• After 6 months, our evaluation and benchmarking setup runs on every change, and you have measurably improved coding quality or reduced token costs per case.
• After 12 months, at least one new approach from research is running in production, and parts of our pipeline run on smaller models without losing quality.
• Throughout, our AI-related product decisions are grounded in your benchmarks, and the team turns to you when a new model or method comes out.

How it works

Just upload your CV with your application. We’ll get in touch with you for the next steps:

• Intro call (30 min): We get to know each other, talk about your background and answer your first questions.
• Technical interview with Christian (CTO) (60 min): We dive into the ML/AI systems you’ve built and how you evaluated them in production.
• Case: You design an agentic approach for a realistic clinical coding problem, including how you would evaluate it, and present it to the team.
• References: We speak with two or three people you’ve worked with, such as former managers or colleagues.
• Meet the team: You spend a day with us on-site in Munich or Berlin, meet our founders Sophia and Christian and see how we work day to day.
• Offer: If it’s a match for both sides, you’ll hear from us quickly.

We look forward to your application, regardless of gender, background, religion, disability, age or sexual identity.

Find more English Speaking Jobs in Germany on Arbeitnow

We show an excerpt and a link, and the full text when the source allows it. Brands and logos shown belong to their owners.

Source: Statistisches Bundesamt (Destatis), table; dl-de/by-2-0; StarIn's calculation

Posted Nov 01, 2026

Posted Nov 01, 2026