SRE Architect – Multi-Cloud Kubernetes, Observability-as-Code & AI-Driven Incident Tooling

I am looking for an expert SRE Architect to elevate infrastructure reliability, observability, and incident response practices. Our platform serves high-scale customer workloads across multi-cloud Kubernetes environments. Your mission will be two-fold: hardening our core observability infrastructure using an Infrastructure-as-Code (IaC) and GitOps approach, and building next-generation, AI-assisted tooling to streamline our incident triage and response. Own the observability stack (metrics, logs, dashboards, alerting) and manage it entirely as code. Infrastructure & Reliability: Own the reliability of our multi-cloud Kubernetes infrastructure. Actively diagnose complex runtime issues involving latency, memory, GPU utilization, concurrency, and model lifecycles. Operations & Runbooks: Author and refine structured, low-context, safe-to-execute runbooks. Oversee incident response, post-mortems, and remediation tracking. Our Tech Stack Orchestration: Kubernetes (Multi-cloud: EKS, GKE, etc.) Observability: VictoriaMetrics / Prometheus, Loki / ELK, Grafana, Alerting pipelines IaC & GitOps: Terraform, Helm, Flux CD / ArgoCD Incident Management: incident.io (or similar) Requirements Proven track record as an SRE Architect or Principal SRE in high-scale, multi-cloud environments. Deep, production-level expertise with Kubernetes, Terraform, Helm, and GitOps workflows. Strong mastery of the Prometheus/Grafana ecosystem managed via code. Experience building automation scripts or tools (Python, Go, or similar) to hook into incident pipelines—bonus points if you have built LLM/AI-assisted tooling for operations. Excellent capability to navigate ambiguity, establish operational processes, and prioritize pragmatism over over-engineering. Note: Prior Machine Learning (ML) infrastructure experience is a plus. A strong curiosity about serving models at scale is plenty.