About the role
You work close to the operating system, and your job is that the person's computer stays dependable exactly when the agent is busiest and they need it most. Scheduling, memory pressure, and the behaviour of the machine under sustained load are the product here, because an agent that makes your laptop unusable is an agent you turn off.
The work
Build process supervision, resource isolation, permissions, storage integration and background execution. Diagnose problems across CPU scheduling, memory pressure and device drivers. Respect the capabilities and restrictions of each supported operating system.
What good looks like
In your first 90 days, ship a supervised local runtime that survives restarts, respects resource budgets and produces useful diagnostics without exposing personal content.
Evidence we look for
Bring strong operating-systems fundamentals and practical experience in Rust, C, C++ or comparable systems tools. Show how you debug concurrency, memory or performance problems rather than listing languages.
What we need to see
- Strong operating-systems fundamentals: scheduling, memory, I/O, and where they interact
- Practical systems programming in Rust, C, C++ or comparable
- You can describe how you debugged a real concurrency, memory or performance problem, in detail
- Instinct for the machine's behaviour under sustained load rather than in a benchmark
Nice to have
- Kernel or driver work
- macOS or Linux internals specifically
- Real-time or latency-sensitive systems
The exercise
Investigate a service that stalls under memory pressure and propose a recovery design that preserves acknowledged work.
Where and how we work
In the office together five days a week, in any of these cities. Remote-friendly around your family, arranged one person at a time.