Senior Site Reliability Engineer (SRE) – CloudVision as a Service (CVaaS)

Who You'll Work With

We’re looking for Site Reliability Engineers to join our growing Arista’s CloudVision-as-a-Service (CVaaS) global SRE team. 

SREs at Arista combine strong software engineering background, systems architecture knowledge, with passion for operating production systems at scale. We are responsible for our global CloudVision service fleet, ensuring scalability, reliability, and stability. You’ll have firsthand experience in being part of a rapidly growing product with a passionate group of engineers that unapologetically put product reliability and customer experience first. We deeply believe in building highly automated and self-sustaining environments, prioritizing safe and efficient operations that leverage cutting edge technologies and tools. 

Arista’s CloudVision is an enterprise network management and streaming telemetry SaaS offering. CloudVision stack is built entirely Kubernetes-native. Familiarity with GCP (Google Cloud Platform) and GKE (Google Kubernetes Engine) is preferred. Our technical stack includes but not limited to: Golang, Python, Ansible/Pulumi, Bash. You will be expected to develop, operate, and work with many different types of databases, both directly on Kubernetes or leveraging managed DB products. We integrate with many different Open Source Software (OSS) projects that both power our microservices stack, monitoring infrastructure, and much more. 

What You'll Do

As an SRE you’ll have the chance to be drive, develop, and lead projects in any of the following areas: 

• Data Platform (NetDL) Architecture and Performance
• Capacity Planning
• Autoscaling
• Disaster Recovery
• Observability
• Change Management - CI/CD
• Service Network Architecture 
• Cost Optimizations
• Instructure and Cloud-First Application Security

You will also be joining globally distributed, “follow the sun model” on-call team where you’ll:

• Continuously improve operational processes by adding automation
• Leading sustainable incident response and blameless postmortems