Senior Devops engineer

Senior DevOps Engineer

The role. Stand up and run the platform's infrastructure — CI/CD, multi-cloud, observability baselines, the daily-deploy pipeline. Two DevOps shared across the org, supporting a real-time platform (voice, desktop, intelligence, AI) where latency and uptime are product features. Startup environment: 1-week sprints, weekly deploys to production, fail fast, move forward.

What you'll own. CI/CD (trunk-based, daily prod deploys, PR-to-staging) · infrastructure-as-code for everything — no hand-built anything · GCP primary, cloud-agnostic deployment templates for multi-cloud · observability and monitoring baselines · uptime as a personal mission · on-call rotation with the SRE · your committed timelines.

Who you are. A ways-to-YES engineer — when a squad needs something, your default is "here's how we do it safely," never a block. Self-starter, grit, show-me mentality: your infra demos too. You love new technology, adapt fast, use AI tools daily to multiply velocity, and consider yourself exceptional. Team player who likes winning.

Requirements

• 7+ years DevOps / platform; deep GCP plus real multi-cloud experience — you've deployed and run production workloads on at least two clouds and built abstractions that keep us portable.
• Strong INFRA-AS-CODE mentality — Terraform-class IaC is how everything exists; if it isn't in code, it isn't real. Reviewable, repeatable, destroyable, rebuildable.
• Has run infra for a daily-deploy, trunk-based shop; CI/CD for high-cadence teams is muscle memory.
• Codes — Go or Python tooling, not just YAML.
• Good networking understanding — protocols and how they actually work: TCP/UDP, TLS, HTTP/2, WebSocket, DNS, load balancing, and ideally RTP/SIP for our media paths. You debug at the packet level when you have to.
• Strong mindset on monitoring and uptime — metrics, logs, traces, alerting baselines from day one; you notice before the customer does.
• GitOps discipline — desired state in git, drift detection, PR-driven infra change.
• Deploy safety engineering — canary/progressive rollout, feature-flag integration, instant rollback; weekly deploys stay boring.
• Kubernetes/container depth (GKE-class) — scheduling, autoscaling, resource tuning for latency-sensitive services.
• Secrets and least-privilege IAM as defaults; SOC 2-ready posture without slowing the train.
• Cost awareness — you catch the runaway bill in the dashboard, not the invoice; efficiency is part of the job.
• Ephemeral environments — spin up a full stack per PR/demo, tear it down after.
• A debugger of infrastructure — reads the deploy log, the metric spike, the routing table, and sees it.