Lead DevOps Engineer

Role overview

Lead DevOps Engineer for Paramount’s Applied Intelligence Personalization team. You will build and maintain scalable, low-latency infrastructure that powers personalization and engagement across Paramount streaming platforms, including real-time ML inference and high-traffic services.

Responsibilities

• Design, implement, and manage scalable Kubernetes-based infrastructure for personalization services
• Build and own CI/CD pipelines for safe production releases (canary, rollback, progressive delivery) including deployment of ML models
• Set up observability and monitoring (Prometheus, New Relic, OpenTelemetry, Grafana), define SLIs/SLOs, and apply error-budget discipline
• Ensure production high availability, security, and performance for APIs and streaming data pipelines
• Partner with application, data, and ML engineers to integrate their workloads into the platform
• Implement autoscaling strategies (HPA, KEDA, traffic-driven) for bursty traffic
• Manage Pub/Sub and event-driven architectures for real-time messaging and engagement analytics
• Optimize hot-path services with caching (Redis, Memcached, etc.)
• Debug and resolve production issues related to latency, scaling, and reliability

Key projects

• Real-time serving infrastructure for personalization and engagement (ML inference workloads)
• Scalable, secure CI/CD pipelines for services and ML models
• Platform-wide log aggregation and monitoring
• Kubernetes serving optimization for minimal latency and efficient resource use
• Improved A/B testing infrastructure for personalized experience measurement
• Enhanced streaming data pipelines for real-time data consumers

Requirements (Basic qualifications)

• 4+ years of production DevOps/SRE/Cloud Infrastructure experience
• Strong Kubernetes and container orchestration experience
• CI/CD experience with GitHub Actions, Jenkins, and ArgoCD
• Deep cloud infrastructure knowledge (GCP, AWS, or Azure)
• Infrastructure as Code (IaC) with Terraform and Helm
• Experience with event-driven architectures/message queues (Pub/Sub, Kafka, etc.)
• Monitoring/logging experience (New Relic, Prometheus, OpenTelemetry, etc.)
• Scripting skills in Python, Bash, or Go for automation
• Proven ownership of production reliability (SLIs/SLOs, incident response, postmortems)

Nice to have

• Experience deploying and operating ML models in production (TF Serving, Triton, TorchServe, Ray Serve)
• GPU/accelerator scheduling and Kubernetes node-pool management
• Knowledge of load balancing, API gateways, and caching at scale
• Experience with A/B testing frameworks and experimentation infrastructure