Senior MLOps Engineer

You will deploy and maintain production ML infrastructure, optimize GPU utilization, and serve large and small language models. You will build CI/CD pipelines, create Helm templates for Kubernetes deployments, implement model optimization and serving workflows, and set up monitoring, logging, and automated workflows to ensure reliable model delivery.

Responsibilities

• Deploy scalable production-ready ML services with optimized infrastructure
• Manage and autoscale Kubernetes clusters
• Optimize GPU resources using MIG and NOS
• Manage cloud storage to ensure high availability and performance
• Integrate LoRA and model merging workflows
• Adapt and deploy state-of-the-art ML codebases
• Deploy and manage LLMs SLMs and LMMs
• Serve models using Triton Inference Server and other serving frameworks
• Leverage vLLM and TGI for model serving
• Optimize models with ONNX and TensorRT
• Develop Retrieval-Augmented Generation systems
• Set up monitoring and logging with Grafana Prometheus Loki Elasticsearch and OpenSearch
• Write and maintain CI/CD pipelines using GitHub Actions
• Create Helm templates for rapid Kubernetes node deployment
• Automate workflows using cron jobs and Airflow DAGs

Requirements

• Bachelor's or Master's degree in Computer Science Engineering or related field
• Proficiency in Kubernetes Helm and containerization technologies
• Experience with GPU optimization including MIG and NOS
• Experience with cloud platforms such as AWS GCP and Azure
• Knowledge of monitoring tools such as Grafana and Prometheus
• Proficiency in scripting languages Python and Bash
• Hands-on experience with CI/CD tools and workflow management systems
• Familiarity with Triton Inference Server ONNX and TensorRT