A strong systems thinker who is equally comfortable leading technical direction and getting hands-on with implementation
7+ years of software engineering experience, with significant time in infrastructure, platform, or ML systems roles
Hands-on reliability engineering experience — you have well-formed convictions about observability, monitoring, deployment systems, and loosely coupled architectures, and you've put them into practice at scale
Proven track record of building and shipping services at scale, with all the operational complexity that comes with it
Kubernetes — significant production experience building, operating, and scaling Kubernetes clusters
Experience designing and shipping flexible domain models and APIs — you think carefully about boundaries, contracts, and long-term maintainability
A default toward automation — you've consistently built efficiency gains through automation and have the track record to show it
Strong communication skills — you can lead your own direction, write clearly about tradeoffs, and bring engineers and stakeholders along with you
We'd love to hear about experience in any of these areas — we don't expect any one person to have all of them
Infrastructure as Code at scale — significant IaC experience, preferably Terraform; CloudFormation, Pulumi, or Kubernetes-based approaches also welcome. Ideally you've led, architected, or contributed to a multi-stack, self-serve IaC system and understand the challenges of building infra that teams can own independently
ML infrastructure — any combination of the following
PyTorch experience, especially model optimization for serving
ML training or serving experience in general
Experience building ML serving and/or training infrastructure (TorchServe, Seldon, KServe, Ray Serve, or similar)
Experience building and leading large-scale distributed training and serving systems
Data engineering — pipeline design, dataset management, or data platform experience
Database design — complex schema design, query optimization, and hard data modeling decisions across relational and non-relational stores
Real-time communication systems — low-latency audio, video, or streaming infrastructure