DevOps Engineers (AIOps) - API and Kubernetes Dicetek LLC
Posted
Summary
DevOps engineer in Abu Dhabi owning the Azure infrastructure for an AI agent platform (AWA): Terraform/Bicep IaC, Azure DevOps CI/CD pipelines, a 5-namespace AKS cluster, container image signing via ACR/Notary v2, Key Vault secrets, Azure Monitor/PagerDuty alerting, security hardening, and disaster recovery testing.
- Infrastructure as code: Own and maintain all AWA Azure infrastructure in Terraform/Bicep: AKS 5-namespace cluster, ADLS Gen2 accounts with WORM policies, Azure AI Foundry APIM configurations, Azure Firewall allowlists, private endpoints, Key Vault, Container Registry, ExpressRoute peering.
- CI/CD pipeline ownership: Design and maintain Azure DevOps pipelines for all AWA components: container image build - ACR push - Notary v2 signing - AKS deployment; Terraform plan - policy check - apply; Agent Regression Testing gate on every PR.
- AKS operations: Manage the 5-namespace AKS cluster - namespace isolation, Network Policies, Pod Disruption Budgets, KEDA autoscaling rules based on Kafka consumer lag, OPA Gatekeeper admission policies, Workload Identity bindings for all agent pods.
- Container governance: Manage ACR, enforce Notary v2 image signing, configure vulnerability scanning, maintain the approved-image ConfigMap used by OPA Gatekeeper admission control.
- Secret and certificate management: Own Azure Key Vault configuration - access policies per managed identity, automatic TLS certificate rotation, secret rotation schedules, audit of all secret access events.
- Monitoring and alerting: Configure Azure Monitor alert rules and Log Analytics KQL queries for SLA breach, error rate spike, pod crash loop, token budget breach, and Kafka consumer lag; maintain PagerDuty escalation paths.
- Security posture: Work with the AI Security Engineer to implement Azure Policy assignments, Defender for Cloud recommendations, Firewall rule reviews, and Private Endpoint configurations; maintain the AWA network topology diagram.
- Disaster recovery: Maintain and test DR procedures: failover to UAE South replica for ADLS Gen2, Azure AI Foundry PTU - PAYG - DR endpoint routing, Orkes cluster recovery, RTO/RPO validation.