Une candidature complète en une minute
Un CV et une lettre de motivation personnalisés, prêts à envoyer.
HalloCharly / Zaitgeist in Berlin is seeking an experienced AI/ML engineer to design and evolve complex RAG systems for voice and chat bots. You will co-own the system architecture, including continuous evaluation and monitoring, and help build secure cloud infrastructure with a strong emphasis on data protection.
You will continually improve the systems based on concrete user feedback, driving fast prototypes to production with reliable, scalable code and modern tooling.
We are a highly specialised tech startup in the heart of Berlin – working on one of the hardest and most relevant problems of our time: bringing artificial intelligence to where it has the biggest leverage – the German state. What we build isn't theory; it's already in use today. Our solutions support more than 1.4 million people across most German federal states – and the number is growing fast. As one of Germany's fastest-growing GovTech startups (50% MoM growth over several months), something genuinely new happens here every day. This is one of those places where hours fly by and weeks feel like months. A lot is happening. We have turned down VC offers so far. So no politics, no theatre – just 100% customer focus. You work directly with the founders (ex-UN, ex-TR, PhDs, German Academic Scholarship Foundation, top-of-class) as part of a small, excellent tech collective with experienced senior advisors. In short: maximum impact, minimum friction. It doesn't get more "streamlined speedboat" than this.
Design and evolve complex RAG systems for voice and chat bots
Co-own the system architecture, including continuous evaluation and monitoring
Build secure cloud infrastructure with a strong focus on data protection
Keep improving the systems based on concrete user feedback
A real appetite for startups (corporate mindsets won't be happy here)
~2 years independently building AI systems (e.g. LLM chatbots, AI-native apps)
~2 years with RAG systems and crawling
~5 years with distributed systems and production operations
The ability to keep a cool head between speed and substance
You build working prototypes fast, use modern coding assistants and LLM tools, and test hypotheses directly on the product
Then you turn those solutions into production-ready code with clean architecture, tests, monitoring and high operational reliability
Nice-to-have: experience in security-critical or regulated environments
Nice-to-have: experience continuously adapting evaluation systems
A beautiful, central office right at Rosenthaler Platz
28 days of vacation per year
Direct influence on technical vision, architecture and company growth
Very exciting growth opportunities