AI/MLOps Engineer with DevOps - 2 No.s

Job Details

Role: AI/MLOps Engineer

Employment Type: 1 year & extendable contract

Work Location: Abu Dhabi, United Arab Emirates

Notice Period: Immediate joiners to a max of 15-30 days of notice

Required Experience:

• Minimum 3+ years in DevOps, MLOps, or AI infrastructure roles.
• Total 10+ years in technical solution operations (systems, infrastructure, or software delivery).
• Proven hands-on administration of local OpenShift or Kubernetes clusters (≥ 1 year) in production.
• Preferably holds CKA / CKAD / OpenShift certification.
• Experience managing multi-node GPU clusters with NVIDIA driver stack, GPU allocation and scheduling (taints/tolerations, device plugin, MIG tuning, etc.)
• Familiarity with GPU workload monitoring and tuning (e.g., DCGM, nvidia-smi, Prometheus).
• Deployed full-stack LLMs (e.g., LLaMA, Mistral, Qwen) on on-prem GPU Kubernetes/OpenShift clusters.
• Proven fine-tuning experience, like using LoRA, QLoRA, PEFT, or full model retraining.
• Managed training/fine-tuning datasets, tokenization, and version control.
• Familiar with deploying PyTorch or TensorFlow models in production-grade inference setups.
• Built and managed end-to-end CI/CD pipelines, with tools like GitLab CI, Jenkins, ArgoCD, or Tekton on K8s/OpenShift.
• Experience integrating model training, packaging, deployment automation.
• Operated production model registry systems like MLFlow Tracking/Registry or KubeFlow Pipelines.
• Able to track experiment metadata, versions, stages, and deployment status.
• Implemented real-time monitoring with Prometheus, Grafana, and GPU exporters (e.g., DCGM).
• Proficient with Linux system administration, Python/Bash scripting, and security hardening of K8s environments.
• Has deployed and operated vector databases (e.g., FAISS, Weaviate, Qdrant, Milvus).
• Integrated with LLM inference pipelines or RAG architectures.
• Any relevant certifications would be a huge plus