Building the runtime platform that gives AI agents persistent, secure environments to act on the real world.
Software Engineer, Infrastructure
We're a small, well-funded team that believes AI agents are only as capable as the infrastructure they run on. Our platform gives autonomous systems durable sandboxes, long-running compute, and the kind of production guardrails that defense and enterprise customers demand before they'll trust a machine to operate unsupervised. This isn't model research — it's the layer underneath that makes agents reliable, observable, and safe. You'll design the control plane that orchestrates thousands of isolated environments across clouds, shaping how the next wave of software gets built and deployed.
What they're looking for
- 4–7 years writing production infrastructure code, with deep fluency in Kubernetes, container runtimes, and at least one major cloud provider (AWS or GCP)
- Comfortable owning the full observability stack — you've built dashboards in Grafana, tuned Prometheus alerts, and know what healthy SLOs look like for distributed systems
- Strong Terraform or equivalent IaC experience; you treat infrastructure as software and have opinions on module design and state management
- Systems-level debugging instincts: you can trace a request from a container's network namespace through a service mesh to a cloud load balancer without guessing
- Clear, low-ego communicator who writes design docs before writing code and defaults to sharing context with a tight-knit, 19-person team