Infrastructure Engineer - Developer Platform

TypeSafe AI•San Francisco, CA

1d•Onsite

About The Position

TypeSafe is a frontier model lab building reliable and general AI systems to power economically valuable automation. Their mission is to usher in a new era of Transformative Artificial Intelligence (TAI). They are looking for an Infrastructure Engineer to build and operate the infrastructure behind TypeSafe AI's products at global scale. This role involves owning systems that serve millions of users across regions, from provisioning Kubernetes clusters across multiple clouds to optimizing networking for low-latency AI inference. It's a high-impact role on a small, fast-moving team, working across the full infrastructure stack: cloud primitives, container orchestration, networking, observability, and specialized infra for large-scale model inference efficiency.

Requirements

Deep experience with Kubernetes in production at scale: networking, storage, scheduling, upgrades
Strong background in AWS
Hands-on experience with infrastructure as code (Pulumi, Terraform, or similar)
Solid understanding of Linux networking
Track record with high-traffic production ML systems
Programming fluency, Python preferred

Nice To Haves

Experience with large-scale LLM / ML inference infrastructure (GPU scheduling, model serving, vLLM, KubeRay, Kubernetes-native tooling)
Kubernetes networking depth with Cilium or other CNI plugins; service mesh (Istio, Envoy)
Multi-cloud infrastructure
Background in site reliability engineering including SLOs, incident response, capacity planning

Responsibilities

Design, deploy, and operate Kubernetes clusters across multiple regions and clouds
Build and maintain infrastructure for the platform that powers LLM inference workloads globally
Own networking, including VPCs, peering, load balancing, DNS, service mesh, CNI
Manage GPU infrastructure and autoscaling for ML workloads
Write and maintain infrastructure as code (Pulumi / Python)
Operate and improve observability: monitoring, alerting, tracing, logging