MLOps Engineer (m/f/d)

Implement and operate CI/CD pipelines, automated testing and release processes for AI/ML workloads. Build and maintain model registry, model serving and AI gateway integrations for LLM APIs and internal applications. Configure and maintain observability for model usage, cost, token consumption, latency, reliability and quality signals using tools such as Prometheus, Grafana, logging and alerting platforms.

Support the transition of workloads from sandbox or PoC environments into production by following defined standards, runbooks and support models. Implement reusable technical components for LLM API integration, RAG pipelines, evaluation pipelines and integration with business applications. Execute infrastructure-as-code for platform environments across container and cloud infrastructure, including Docker, Kubernetes and Helm-based deployment patterns.

Maintain runbooks, operating procedures, technical documentation and operational dashboards for platform components. Support incident analysis, reliability improvements, cost optimization and lifecycle maintenance for production AI workloads. Work with nearshore, system integration or cloud partners on specific implementation tasks as directed by the AI Platform Engineer.

Collaborate with data engineering, application development, cloud platform and security teams on integration, identity, access and deployment requirements. 3-5 years of experience in DevOps, cloud engineering, ML engineering, MLOps or platform engineering. Hands-on experience with CI/CD, infrastructure as code and automated deployment in production environments. Strong practical Python skills and Git-based development workflows. Experience with Docker and Kubernetes; deployment tooling such as Helm is desirable.

Working experience with Azure or AWS cloud services, including compute, storage and IAM concepts. Experience with observability tooling such as Prometheus, Grafana, logging platforms and alerting practices. Familiarity with MLOps concepts such as model registries, evaluation pipelines, drift monitoring and model lifecycle management. Understanding of network isolation, identity, secrets management and API access control. Fluency in English, spoken and written.