AI/ML DevOps Engineer

NTT DATA is seeking a skilled AI/ML DevOps Engineer to join their team in Noida, Chennai, Bengaluru, Pune, or Uttar Pradesh, India. This hands-on role is crucial for designing, building, and maintaining Infrastructure-as-Code (IaC) templates using Terraform to provision and manage Azure AI/ML services for various application teams. The position requires a blend of engineering expertise (Terraform, CI/CD pipelines, Azure Kubernetes Service - AKS, security) and coordination skills (Kanban methodology, cross-team alignment, risk management) to facilitate secure, repeatable deployments of AI workloads onto the bank's Azure AI Platform.

Key responsibilities include:

• Developing IaC modules and reusable Terraform templates for provisioning Azure resources essential for AI/ML workloads, such as Azure OpenAI/AI Studio, Azure ML, Cognitive Search, Key Vault, Storage, and networking components.
• Automating deployment pipelines using Azure DevOps or GitHub Actions, incorporating plan/apply steps, policy validations, and environment promotions, along with integrations for secrets management, approvals, and drift detection.
• Implementing access and identity management using Microsoft Entra ID patterns (app registrations, group/role assignments, RBAC) for application teams and automation processes, with a specific focus on understanding security related to Entra IDs.
• Supporting deployments on AKS and integrating platform components (ingress, images, namespaces, quotas) for AI services hosted on Kubernetes.
• Enhancing security and governance by embedding guardrails through Azure Policy, role assignments, private endpoints, and establishing baselines for tagging, FinOps, logging, and monitoring.
• Utilizing Python or Bash scripting for IaC tooling and deployment utilities.
• Experience in codifying policies and controls (e.g., OPA/Conftest, Azure Policy as Code) and implementing cost governance tags is also required.

This role is ideal for an engineer passionate about building robust AI infrastructure and ensuring secure, efficient deployments.