Machine Learning Engineer

Repon GmbH; Aachen; Vor Ort; EUR 70.000 - 110.000; Vollzeit

Benefits dieser Stelle

Flexible working hours

Home office by arrangement

Zusammenfassung

Repon GmbH, based in Aachen, is seeking a machine learning engineer to own ML problems end-to-end—from conception through deployment and ongoing improvement. You will work closely with a small team to apply ML across computer vision, NLP, and document understanding, delivering measurable results in enterprise projects.

The role emphasizes hands-on development, collaboration with backend and frontend teams, and clear technical communication in English.

Qualifikationen

• 2+ years in production machine learning engineering.
• Proficiency in Python and ML frameworks such as PyTorch or TensorFlow.
• Experience building ML services with FastAPI and async patterns.

Aufgaben

• Own ML solutions end to end from concept to deployment.
• Explore, evaluate, and iterate on models and data.
• Collaborate with backend/frontend teams to ship features.
• Design systems with user focus and clear benchmarks.

Kenntnisse

ML ownership

Python

ML frameworks

PyTorch

TensorFlow

FastAPI

Tech comm

English

Tools

Docker

GitHub Copilot

Claude Code

CI/CD

Jobbeschreibung

This is not a remote position. We work primarily together from our office in Aachen-Burtscheid. Occasional home office is possible by arrangement, but regular on-site presence in Aachen is an essential part of the role.

About Us

Repon GmbH is an innovative technology company specializing in AI-driven process optimization for businesses. We develop and integrate tailored solutions that help organizations automate complex processes using both structured and unstructured data.

Our strength lies in a pragmatic, human-centered approach, with a strong focus on clarity, reliability, and measurable results. We work on challenging enterprise projects for clients such as Deutsche Telekom, BSH, and leading companies in the automotive and financial sectors.

As a small, dynamic team with direct access to decision-makers, we offer a genuine startup environment, fast decision-making, and plenty of freedom to contribute your own ideas. At our office in Aachen-Burtscheid, you can expect an open and collaborative company culture with a strong emphasis on creating a comfortable and enjoyable working environment.

Your Role & Responsibilities

• Own ML solutions end to end. Take responsibility for ML problems from the initial requirement or idea through exploration, implementation, evaluation, deployment, and ongoing improvement. You'll have substantial ownership over technical decisions and what ultimately ships.
• Find the right approach. Understand the problem, explore possible solutions, run experiments, and make informed trade-offs based on quality, complexity, performance, cost, and maintainability. The solution is not predetermined: it might involve training or fine-tuning a model, using existing models or services, combining different approaches, or applying deterministic or statistical methods.
• Work across different ML domains. Our work spans computer vision, NLP, document understanding, multimodal ML, and extracting value from unstructured data. You should be comfortable moving between domains rather than being limited to a single specialization.
• Build what the solution needs. When appropriate, develop datasets, annotation and human-review workflows, evaluation systems, post-processing, confidence logic, geometric reasoning, or other supporting components.
• Define what "good enough" means. Design benchmarks, metrics, test sets, and release criteria that reflect real product requirements and allow us to make evidence-based decisions about what ships.
• Put solutions into production. Deploy and operate ML systems, monitor their behavior, investigate failures, and improve them based on real-world performance.
• Integrate ML into the product. Collaborate with backend and frontend teams to turn ML capabilities into complete product features and design clean interfaces between ML components and the rest of the platform.
• Design Systems with the User in mind. When designing systems we always think about the end user experience first, not the technical background.

What we’re looking for

• 2+ years in production machine learning engineering, with real ownership of ML systems from development through deployment, operation, and iteration.
• Proficiency in Python and ML frameworks such as PyTorch or TensorFlow
• Strong ML fundamentals - the concepts behind the work, not just the APIs: Strong fundamentals behind backbones, NLP and Computer Vision tasks, metrics and approaches, and optimization.
• Excellent problem-solving ability. You can take an ambiguous problem, structure it, investigate different approaches, and make sound technical decisions rather than automatically reaching for a familiar tool.
• Backend development for ML systems. You have hands-on experience with FastAPI and asynchronous programming patterns and can build reliable services around ML capabilities.
• Clear technical communication. You can communicate approaches, set backs, timelines and benchmarks to both a technical and a not technical audience.
• Professional working English. You collaborate, review, and document effectively in English.

Nice to have

• Experience with compliance-sensitive or auditable ML systems.
• Experience designing human-in-the-loop workflows for review, annotation, or feedback.
• Experience working with large-scale and different types of unstructured data.
• Applied research experience or a track record of quickly evaluating and adopting current ML techniques.
• Interest in current SoTA. ML is developing fast. Reading papers, understanding implementations and a critical mindset.
• Strong software engineering practices. You have experience with Docker, CI/CD, automated testing, and building maintainable production systems as well as feeling comfortable working with large code bases.
• AI-assisted development - you use modern AI coding tools such as GitHub Copilot, Codex, or Claude Code as part of your workflow.
• German - helpful for team and customer context, but not required.

What we offer

Team & Culture:

• Short decision-making paths - a small team with direct communication and little bureaucracy.
• Technical discussions as equals - you're encouraged to challenge architectural decisions and actively shape how we build things.
• Varied and interesting work - different applications and technical challenges rather than repetitive routines.

How we work:

• Flexible working hours - need to see a doctor at 10? No problem.
• Home office by arrangement - typically around 20%, with flexibility for individual circumstances (This is intended as flexibility within an office-based role, not as a remote-working model.)
• Modern technology - we keep our stack current and regularly update rather than accumulating legacy technology.
• AI-first tooling - premium access to tools such as ChatGPT, Claude, and Claude Code.

Compensation & Development:

• Competitive salary - based on experience.
• Professional development - conferences, courses, or books when they support your work and development.