JOB-1439
An early-stage infrastructure startup is building long-lived, isolated computing environments where AI agents can run software and keep their state over time. It is hiring a systems engineer to work on the core platform, directly alongside the CTO and a small group of engineers, with plenty of freedom to choose an approach.
Areas of work
- Own the control plane: the interfaces and backend services that provision, place and manage the environments, with careful handling of concurrency, persistent state and failure recovery
- Improve the virtual machine monitor and runtime, which are written in Rust, including guest memory handling, device emulation, Linux I/O, snapshots and startup behavior
- Profile and cut performance costs in request and startup latency, kernel call overhead, memory allocation, scheduler behavior, and network and disk I/O, backing each change with baselines and measurements
- Design secure interfaces with explicit trust boundaries, and find security weaknesses with a demonstrated way they could fail or be attacked
- Carry projects from an initial experiment to a verified, deployed and operated service
Who fits
- Strong Linux internals knowledge and the curiosity to dig beneath an API
- Depth in either control-plane services (Go, Rust or similar) or practical Rust plus low-level systems work; spanning both is welcome
- A track record of picking things up quickly and delivering, with recent projects you can walk through in detail
- Security-minded precision about real versus speculative risk
- Clear written and spoken communication, including design notes and code reviews
- Responsible use of contemporary AI coding assistants on top of solid fundamentals
- Any career stage is acceptable; demonstrated ability matters more than years or pedigree
On-site in the San Francisco Bay Area. Visa transfers and new sponsorships are both considered.
