MLOps / Cloud Deployment Engineer

Our Client's Digital Finance IT is scaling AI and agentic systems in production. We need an MLOps / Cloud Deployment Engineer to own the deployment, reliability, observability, and operational scale of these systems in a regulated enterprise environment.

This is a cloud and platform engineering role with deep MLOps/LLMOps focus, not a model-building role. You will operate the runway that ML and GenAI systems run on, not build the models themselves.

What You'll Do

• Own CI/CD pipelines for ML models, RAG applications, and agentic AI systems — from experiment to production
• Deploy and operate AI workloads on cloud-native ML/AI platforms — AWS Bedrock/SageMaker, Azure AI Foundry / Azure Machine Learning, or equivalent
• Build and maintain observability, tracing, and monitoring for LLM and agentic systems — latency, cost, hallucination rates, tool-call success, drift detection
• Implement model governance and guardrails — approval gates, kill-switches, escalation paths, audit trails
• Manage infrastructure-as-code (Terraform, Bicep, or equivalent) for reproducible AI/ML environments
• Design cost and performance optimization strategies — token usage tracking, caching, model routing, autoscaling, warehouse/cluster right-sizing
• Own security posture — RBAC, secret management (Key Vault / Secrets Manager), prompt-injection risk mitigation, auditability for regulated pharma
• Partner with data engineers, AI engineers, and Finance business stakeholders to move systems from prototype to reliable production
• Implement evaluation frameworks for AI systems in production — regression testing, adversarial testing, accuracy tracking, hallucination monitoring

Requirements

Must-Have Experience

• 5+ years in cloud/DevOps/MLOps engineering on AWS, Azure, or GCP
• Production deployment of ML or GenAI systems — CI/CD, containerization (Docker/Kubernetes), infrastructure-as-code (Terraform)
• MLOps tooling — MLflow, SageMaker Pipelines, Azure ML Pipelines, or equivalent
• LLM/GenAI operational experience — observability tools (LangSmith, Weights & Biases, or equivalent), cost monitoring, latency optimization, prompt/model versioning
• Cloud-native AI platforms — hands-on with at least one of: AWS Bedrock, SageMaker, Azure AI Foundry, Azure OpenAI, Vertex AI
• Python, Bash, and infrastructure scripting — strong
• Security and governance in regulated environments — RBAC, secrets, audit, compliance

Nice to Have

• Pharma, life sciences, or regulated financial services domain
• Experience operating agentic AI systems in production — multi-agent orchestration, tool-calling, human-in-the-loop workflows
• LangChain, LangGraph, CrewAI, AutoGen, or Semantic Kernel operational experience
• Kubernetes-native ML platforms (Kubeflow, Ray)
• Snowflake or Databricks operational experience (compute governance, cost management)
• Certifications: AWS/Azure ML Engineer, Kubernetes CKA/CKAD, Terraform Associate

What We're NOT Looking For

• Data Scientists or research engineers — this is a production platform role
• Application developers with light DevOps exposure — need real MLOps/cloud engineering depth
• Pure infra engineers with no AI/ML operational experience — need to understand what makes LLM systems different (evals, hallucinations, prompt versioning, RAG grounding)