Tick the boxes if you agree: ⬜ You love managing your own projects and decisions, from ideation to deployment. ⬜ You know how to talk to both LLMs and humans, aligning AI solutions with real-world product needs and stakeholder expectations. ⬜ You’ve shipped end-to-end ML pipelines, fine-tuned BERT models, and set up MLOps workflows that scale. Our Mission Ever built something with LLMs and thought, “this could actually solve a real problem”? We’re looking for someone who blends curiosity with execution, someone who doesn't just explore NLP and MLOps trends but knows how to translate them into scalable, valuable products. In this role, you’ll shape our machine learning strategy hands-on, own projects end-to-end, and work closely with product and engineering to bring user-first ML solutions to life. Autonomy, impact, and learning included. That sounds like you? Let's talk! At how.fm, we are a growing and experienced team of long-time colleagues. We have made it our purpose to give everyone a fair chance to learn, no matter their background. To achieve this, we have made it our mission to create a digital coach that revolutionizes training for companies and their employees in up to 30 languages. We have the opportunity to meet at the ‘Startplatz’ in Cologne, but we mainly work remotely and encourage everyone to work where it suits them best. At least once every trimester, we have workshops or events where we meet as a team in Cologne. So even if we don't have to commute every day, you should be able to do so regularly. Your everyday activities: Manage and deliver NLP/LLM projects from start to finish Design MLOps/LLMOps infrastructure and workflows (e.g., Langfuse) Explore and integrate third-party ML tools and LLM models (e.g., mlflow, HuggingFace, OpenAI, Gemini) Host internal workshops and syncs with stakeholders Collaborate with Product leads on project scoping and roadmaps Stay sharp on emerging NLP trends and best practices You will need: 3–5 years hands-on in ML/NLP, with a portfolio of production-level projects Deep understanding of modern NLP (transformers, embeddings, context engineering) LLM integration experience (OpenAI, Claude, HuggingFace, etc.) Proficiency in Python and tools like transformers, FastAPI, PyTorch, and sklearn Ability to navigate ambiguity, structure your work, and communicate impact to non-technical teams Experience working in a lean startup environment A Bachelor’s or Master’s in a relevant field, or equivalent real-world proof About us You are a how.fm fit and a fit for this role if you identify with the following: We are humble and goal-oriented: While there is much we have yet to discover, our natural curiosity and passion for learning drive us to continuously assess and enhance the effectiveness of our processes and initiatives. We work with a lean approach on a path guided by the Kaizen philosophy. We make a plan, try it out early, fail, improve it, and eliminate waste. We embrace change and adapt when necessary to stay aligned with our mission. We are always there for each other, like in a tribe, we know that together we are better. If you want to know more about our team and how we work visit: www.how.fm/tribe. If you like what you read and want to join us or want to apply and ask us more about the position, go ahead! We are looking forward to hearing from you! :) We are deeply committed to fostering an inclusive and diverse environment and welcome applications from individuals of all genders, nationalities, ethnic and social backgrounds, religions or worldviews, abilities, ages, and sexual orientations.