Senior Software Engineer, Inference Cloud

About the role

Cerebras Systems builds the world's largest AI chip, delivering training and inference speeds over 10 times faster than GPU-based cloud services. The Inference Cloud team owns the cloud layer behind the Inference Service, covering availability, latency, reliability and global scale. This is a hands-on individual contributor role focused on building cloud services and the Kubernetes infrastructure that runs them, taking projects from design through rollout and ongoing operation.

Key responsibilities

• Kubernetes infrastructure: build and improve production environments on Amazon EKS, including workload scheduling, autoscaling, cluster upgrades, resource utilization, network policies and failure recovery.
• Service networking and traffic control: develop and operate traffic management with Envoy and Istio, improving routing, rate limiting, connection management, timeouts, retries and failover.
• Databases: build database-backed services, troubleshoot query performance, connection pooling and availability issues, and strengthen backup and recovery.
• Reliability and performance: identify bottlenecks and improve latency, throughput and capacity in high-QPS systems, including during traffic spikes and partial failures.
• Infrastructure and deployment automation: use Terraform, Argo CD and Kustomize to automate provisioning and delivery, and improve deployment validation and rollback.
• Production operations: build metrics, logs, traces, dashboards and alerts, investigate incidents and contribute to design and code reviews.

Required profile

• 5+ years of software engineering experience, including building and operating production services or cloud infrastructure.
• Strong programming skills in Go or another backend or systems language, able to write, test and maintain production software.
• Ability to independently deliver projects, explain technical tradeoffs and collaborate across engineering teams.

Required skills

• Strong Kubernetes expertise, including workload lifecycle, networking, service discovery, storage, autoscaling and troubleshooting.
• Hands-on experience with Amazon EKS or a comparable managed Kubernetes service.
• Experience with Envoy and Istio, or comparable proxy and service mesh technologies.
• Solid understanding of HTTP, gRPC, DNS, TLS and mTLS.
• Experience with production databases, query optimization, indexing and connection management.
• Infrastructure as code and automated deployment with Terraform, Argo CD and Kustomize.
• Preferred: AWS networking, distributed systems design, C++, high-QPS optimization, TTFT and tail latency reduction, ML inference infrastructure.