MLOps Consolidation and Deployment Enablement Expert (m/f/d)

Sie suchen spannende Aufgaben, Raum für Ihre Ideen und ein mitarbeiterfreundliches Arbeitsumfeld?

Dann sind Sie bei uns genau richtig!

Im Rahmen der Arbeitnehmerüberlassung suchen wir für unseren Kunden Roche Diagnostics GmbH ab dem nächstmöglichen Zeitpunkt für 6 Monate einen MLOps Consolidation and Deployment Enablement Expert (m/f/d) in Vollzeit.

Ihre Aufgaben:

Support the initial consolidation phase of the MLOps landscape within the department.

• Primary Focus: Standardize and enable the deployment of machine learning models from the newly consolidated model registry to a variety of internal deployment targets and frameworks in a secure and scalable way
• Deployment Pipeline Development: Design, implement, and document robust, repeatable pipelines for deploying models stored in the consolidated model registry
• Deployment Target Integration: Establish and ensure connectivity and deployment capabilities to various target environments, including Orbital Pipelines and Kubeflow/KServe
• Framework Adaptation: Develop and maintain necessary wrappers, SDK extensions, and configuration templates to ensure models (potentially in various formats/frameworks like TensorFlow, PyTorch, Scikit-learn) can be correctly packaged and served by the target deployment frameworks/environments
• Standardization: Collaborate with the core AIE MLOps team and other stakeholders in the organization to define best practices, tooling, and standards for model serving configurations, monitoring hooks, and endpoint health checks
• Documentation & Knowledge Transfer: Create comprehensive technical documentation, runbooks, and provide training/knowledge transfer sessions to the internal MLOps team for long-term ownership and maintenance of the deployed solutions
• Troubleshooting: Address and resolve technical challenges related to model packaging, deployment failures, latency issues, and framework compatibility across diverse environments.

Das bringen Sie mit:

• Bachelor or Master degree in a relevant field
• At least 8 years of relevant experience
• Very good command of English
• Hands-on experience working with model registry (preferably the AI developer platform Weights and Biases)
• Advanced proficiency in Python. Familiarity with shell scripting
• Expert proficiency in Docker and container orchestration, preferably Kubernetes
• Proven ability to implement and optimize high-availability model serving solutions (REST APIs, gRPC).
• Deep hands-on experience with: Kubeflow, KServe, Dagster
• Strong experience with CI/CD tools (e.g., Jenkins, GitLab CI, ArgoCD) and pipeline automation
• Experience working within cloud environments (preferably AWS) and infrastructure-as-code principles (e.g., Terraform) is a plus