Sr Software Engineer, MLOps

Sourced from justjobs.com

NRG Energy (Vivint Smart Home) is hiring a Senior MLOps Engineer to build the model lifecycle, deployment, observability, and infrastructure foundations behind production AI features such as recognition, AI Video Search, Multimodal AI, Agentic AI, and Energy AI. The role focuses on shared, reliable platform primitives that help AI teams ship faster. You will partner with engineering, data science, product, and operations teams to productionize AI capabilities at scale.

• Build model registry, model serving, deployment, rollback, and CI/CD systems for production AI services
• Own feature, dataset, model, and prompt versioning patterns across AI products
• Standardize training, evaluation, release, monitoring, and operational workflows for AI teams
• Improve reliability, cost efficiency, latency, and repeatability of AI launches
• Create reusable platform patterns across AI features
• Partner with engineering, data science, product, and operations teams to productionize AI at scale

• Bachelor's in CS, Software Engineering, AI/ML, or related field with 5+ years of professional experience; or Master's with 2+ years
• Experience building production ML platforms, model serving systems, or MLOps workflows
• Strong Python and cloud engineering skills
• Experience with CI/CD, Git, infrastructure-as-code, and production monitoring
• Familiarity with model registry, feature/data versioning (DVC), validation, deployment, rollback, and observability
• Ability to communicate tradeoffs clearly across engineering, data science, and product teams

• Experience with GCP/AWS, Cloud Run, Kubernetes, Vertex AI, SageMaker, MLflow, or equivalent tools
• Experience with AI services for computer vision, LLMs, multimodal models, or recommendation systems
• Experience with data validation, dataset versioning, feature stores, or model quality monitoring
• Experience optimizing cost, latency, reliability, and operational readiness for AI systems
• Experience with IoT, edge AI, smart home, or distributed device environments