Senior MLOps Engineer

How you will benefit 

• Salary which entirely fits your expectations
• Full relocation package to Dubai which will be adopted to your specific needs
• Official employment in the UAE

We are seeking a Senior MLops Engineer with proven experience in deploying and managing large-scale ML infrastructure for LLMs, TTS, STT, Stable Diffusion, and other GPU-intensive models in production. You will lead the design and operation of cost-efficient, high-availability, and high-performance serving stacks in a Kubernetes-based AWS environment.

How You Will Influence the Workflow

• You will architect, deploy, and maintain scalable ML infrastructure on AWS EKS using Terraform and Helm.
• You will own end-to-end model deployment pipelines for LLMs, diffusion models (LDM/Stable Diffusion), and other generative/AI models requiring high GPU throughput.
• You will design cost-effective, auto-scaling serving systems using tools like Triton Inference Server, vLLM, Ray Serve, or similar model-serving frameworks.
• You will build and maintain CI/CD pipelines integrating the ML model lifecycle (training → validation → packaging → deployment).
• You will optimize GPU resource utilization and implement job orchestration with frameworks like KServe, Kubeflow, or custom workloads on EKS.
• You will deploy and manage FluxCD (or ArgoCD) for GitOps-based deployment and environment promotion.
• You will implement robust monitoring, logging, and alerting for model health and infrastructure performance (Prometheus, Grafana, Loki).
• You will collaborate closely with ML Engineers and Software Engineers to ensure smooth integration, observability, and feedback loops.

What Will Help You Stand Out

• Experience with model serving frameworks such as Triton, vLLM, Ray Serve, TorchServe, or similar will highlight your expertise in serving AI models.
• Experience deploying and optimizing LLMs and LDMs (e.g., Stable Diffusion) under high load with GPU-aware scaling will set you apart as an expert in handling resource-intensive models.
• Strong expertise in Kubernetes (EKS) and infrastructure-as-code (Terraform, Helm) will showcase your ability to build high-quality, efficient infrastructure.
• Hands-on experience in Python and the full ML model lifecycle in production environments will prove your practical experience in real-world deployments.
• A willingness to explore new languages such as Go or Rust for backend systems or performance-critical systems will give you a competitive edge.

Required Experience

• 2–3 years of experience with model serving frameworks such as Triton, vLLM, Ray Serve, TorchServe, or similar.
• 2–3 years of experience deploying and optimizing LLMs and LDMs (e.g., Stable Diffusion) under high load with GPU-aware scaling.
• 3–4 years of experience with Kubernetes (EKS) and infrastructure-as-code (Terraform, Helm).
• 4–5 years of hands-on software engineering experience in Python, with production-grade experience in ML model lifecycle.
• Nice to have: familiarity with Go or Rust for backend or performance-critical systems.
• Fluent English

Nice to Have

• Experience with model quantization, distillation, and optimization.
• Familiarity with ML model registries like MLflow or DVC.
• Exposure to Kafka or event-driven data pipelines.
• Contributions to open-source MLOps tools or frameworks.