We are seeking a Senior Software Engineer (MLOps) to bring machine learning into production at scale. You will build ML and LLM pipelines, standardize model lifecycle management and help teams ship reliable AI features using Python, container platforms and cloud services.
Responsibilities
- Build and maintain end-to-end ML and LLM pipelines from ingestion to training, serving and CI/CD
- Own model lifecycle workflows including experimentation, registry, deployments, promotions and monitoring
- Operationalize large language models, embeddings, Retrieval-Augmented Generation (RAG) and agentic workflows
- Refactor research prototypes into production-grade Python with testing, debugging and performance tuning
- Containerize services using Docker and orchestrate workloads with Kubernetes
- Define observability for AI systems including drift, quality metrics and reliability signals
- Partner with data science and platform teams to integrate AI capabilities into shared services
- Contribute to engineering standards for security, compliance and reproducibility
Requirements
- Strong background in Python with the ability to write clean, testable, maintainable code
- Experience building production ML systems including CI/CD and release workflows
- Hands-on experience with Docker and Kubernetes for packaging and running workloads
- Familiarity with experiment tracking and model registry tools such as MLflow or Weights & Biases
- Proven experience deploying and operating large language models including embeddings and RAG patterns
- Knowledge of monitoring approaches for ML systems including drift and performance tracking
- Understanding of cloud compute and GPU fundamentals for training and inference optimization
- Ability to collaborate across data science, engineering and infrastructure teams
Nice to have
- Experience with LangChain and LangGraph or similar orchestration frameworks
- Experience with GitLab CI or Jenkins in production environments
- Interest in keeping up with applied ML and AI advances and translating them into shippable features