The posting
Aufgaben Own the architecture and technical direction of production-grade cloud, MLOps and data infrastructure and core capabilities that enable R&D teams across Hexagon to develop and deploy machine learning solutions for robotics Advance the strategy and standards for data ingestion, storage, management and sharing to enable the creation and usage of high-value robotics datasets Design and develop scalable data pipelines to ingest, process and store both real-world and simulated robotics datasets, and facilitate data sharing and usage Design and develop scalable and reproducible ML pipelines to train, evaluate and deploy ML models for robotics Own the cloud and hybrid foundations underneath: compute and GPU scheduling, storage, networking, identity and access management, observability and cost governance Balance the trade-offs between functionality, quality and cost within a given timeline, and make those trade-offs explicit to engineering leadership Assess and select third-party frameworks, platforms and vendors with long-term operational and maintenance considerations in mind, and take part in negotiating terms with external partners Create pipeline templates and work with R&D teams to facilitate the adoption and use of MLOps, cloud and data infrastructure and tooling, write clear documentation, provide hands-on guidance and share knowledge Establish and promote engineering and CI/CD best practices for data and ML pipelines, ensuring versioning, modularity, testability and high-quality releases Work closely with R&D teams to understand their requirements, translating needs into high-quality and effective infrastructure and tooling, and mentor engineers to raise the level of platform engineering across teams Take responsibility for intellectual property in your area of expertise
Anforderungen MSc/PhD in Computer Science or equivalent, or equivalent depth acquired in practice Typically 10+ years of relevant experience, with a T-shaped profile: broad engineering breadth and recognized depth in cloud infrastructure and MLOps Deep understanding of MLOps and data infrastructure concepts and tooling, including the data and ML lifecycle, considering both functional and non-functional requirements Proven track record in architecting, implementing and operating production-grade cloud, MLOps and data infrastructure, with capabilities for data storage and processing, ML experimentation, data and ML pipelines, tracking, registering and deployment of ML models Strong cloud architecture fundamentals independent of any single provider: identity and access management, networking, storage and compute primitives, managed data and ML services, and the cost implications of each. Depth in Azure data and ML services (Azure Data Lake, Azure ML) is a plus; equivalent depth in AWS or GCP is equally welcome Hands-on experience with cloud, hybrid and on-premises environments, including moving large datasets and models between them Solid experience in implementing modular, testable and scalable data and ML pipelines Solid software engineering skills in Python, ability to write clean, testable and maintainable production-ready code, experience with CI/CD, IaC (Terraform), containerization (Docker) and container orchestration (Kubernetes) at production scale Working proficiency in ML frameworks, e.g. PyTorch, TensorFlow, or JAX, and tooling, e.g. MLflow - enough to be a credible partner to the teams building the models Experience in ML models for robotics and in deploying ML-based models to robotic hardware is a plus. Knowledge of robotics frameworks and tools (particularly ROS 2, Omniverse) and the ability to work closely with robotics teams is a plus Strong agile mindset, ability to operate with autonomy, strong communication skills, ability to work in a customer-oriented way with technical and non-technical stakeholders, ability to enable teams Proficiency in English required, basic German skills advantageous
Sprachen Proficiency in English required, basic German skills advantageous
