About the Role
As a working student, you will support our team in connecting and further developing the various software, infrastructure, and ML components of our research and development platform into a functioning overall system. The platform monitors the security and robustness of machine learning systems throughout their entire lifecycle – as a cloud-native system based on Kubernetes, which we are currently transferring to an OpenStack-based cloud infrastructure. We work according to modern development and operating principles (Git-centric, IaC, containerization, CI/CD).
Responsibilities
- Collaborate on our layered IaC architecture with OpenTofu/Terraform (modular layers for cloud infrastructure, platform services, and applications)
- Provision and maintain Kubernetes environments, prospectively on an OpenStack-based cloud, including network, storage, and access configuration
- Deploy components using Helm, Kustomize, and K8s manifests
- Troubleshoot cluster issues (Pods, Services, Logs, Events) using kubectl
- Build and maintain Docker images for our services (multi-stage builds, build automation)
- Manage images in our internal container registry
- Operate and integrate MLOps components: Kubeflow Pipelines & Training Operator, MLflow (Experiment Tracking & Model Registry), KServe (Model Serving), MinIO (S3-compatible storage)
- Support the registration, deployment, and versioning of ML models, as well as setting up training and serving pipelines
- Connect and maintain the streaming layer for inference logging (Kafka-compatible message broker, CloudEvents)
- Implement and configure interfaces between services (REST, GraphQL, message queues, model inference protocols)
- Contribute to our ML services in Python, e.g., for model monitoring, drift detection, and the demo application
- Support evaluations, diagnostic plots, and method comparisons
- Create and execute unit, integration, and smoke/end-to-end tests
- Automate recurring deployment and build steps; contribute to building CI/CD pipelines
- Maintain architecture, deployment, and runbook documentation (Markdown)
- Prepare demo, presentation, and reproduction materials
Requirements
- Enrolled student (Computer Science, Data Science, Computational Engineering, Electrical/Communications Engineering, or comparable)
- Solid Python skills and enjoyment in reading and understanding foreign code
- Basic understanding of containers (Docker) and Kubernetes
- Proficient use of Git and the command line (Linux)
- Independent, solution- and process-oriented work style, as well as willingness to quickly learn new technologies
- Good German and English language skills, both written and spoken
- Experience with Infrastructure-as-Code (Terraform/OpenTofu)
- Experience with cloud infrastructure, ideally OpenStack
- Knowledge in the MLOps environment (MLflow, Kubeflow, KServe, MinIO)
- Experience with Kafka or event-driven architectures
- Experience with ML frameworks (e.g., PyTorch) is advantageous
- Interest in ML security & robustness (attacks on models, model monitoring)
- Initial experience with agile methods (Scrum, Kanban)
Skills
- Python
- Docker
- Kubernetes
- Git
- Linux command line
- OpenTofu
- Terraform
- OpenStack
- MLflow
- Kubeflow
- KServe
- MinIO
- Kafka
- PyTorch
- REST
- GraphQL
- Message queues
- CloudEvents
- Helm
- Kustomize
- kubectl
- Markdown
Location
- Paderborn
- Hamburg
- Berlin
Work Type
- Hybrid
- Remote possibility
Experience Level
- Working student
Education Level
- Student (Informatik, Data Science, Computational Engineering, Elektro-/Nachrichtentechnik oder vergleichbar)
Benefits
- Practical involvement in a real research and demonstrator project at the intersection of MLOps, cloud-native infrastructure, and ML security
- Deep insights into a modern, consistent technology stack (OpenTofu, OpenStack, Kubernetes, Kubeflow, MLflow, KServe, Kafka)
- Independent tasks with direct, visible contribution to the system
- Close supervision and mentoring by experienced engineers – short paths, direct feedback, and room to take responsibility
- Flexible, study-friendly working hours
- Prospect of long-term collaboration after graduation
About the Company
- The Atos Group is a global leader in digital transformation.
- With approximately 56,000 employees and an annual revenue of approximately EUR 7.2 billion (based on the future company structure), the company operates in 54 countries under two brands: Atos for Services and Eviden for Products and Systems.
- As the European number one in cybersecurity and cloud, Atos Group works towards a secure and decarbonized future, offering customized AI-powered end-to-end solutions for all industries.
- Atos Group is listed on Euronext Paris.
Equal Opportunity
- At Atos, diversity and inclusion are embedded in our DNA. We welcome your application regardless of origin, religion, color, gender, age, disability, or sexual orientation.
- All decisions throughout the entire recruitment process are based exclusively on qualifications, skills, knowledge, and experience, as well as relevant business requirements.
- We value equal opportunities and welcome applications from people with disabilities.
- In case of equal qualifications, severely disabled applicants and persons with equivalent status are given preferential consideration.
- Read more about our commitment to a fair working environment for all.
- Atos is a recognized market leader in its industry regarding Environmental, Social, and Governance (ESG) criteria.
- Learn more about our CSR commitment.
