Low-Latency Infrastructure at Modal
Modal executes user functions across thousands of bare-metal GPU servers with cold starts measured in tens of milliseconds. Speed matters here. Standard Kubernetes or containerd daemons introduce seconds of boot latency. Modal replaces them with custom Rust daemons that talk directly to Linux kernel primitives. Worker daemons configure cgroups v2, network namespaces, and seccomp filters on the fly. Root filesystems stream lazily from remote blob tiers through user-space block devices. No waiting on disk unpacking. As Systems Engineering Manager, you guide engineers building container runtimes, microVM storage drivers, and gRPC scheduling fabrics. You will bridge kernel low-level internals with high-level team leadership.
Base compensation ranges from $275,000 to $330,000 USD plus equity ownership in Modal.
Core Production Responsibilities
- Lead a focused team of systems programmers writing production Rust on bare-metal Linux.
- Architect container execution environments supporting dynamic NVIDIA GPU allocation and host isolation.
- Prioritize runtime initiatives including user-space filesystem drivers, snapshot restore engines, and memory reclamation.
- Partner with platform engineers to diagnose tail latency regressions across thousands of concurrent tenant tasks.
Technical Background
- Engineering management experience guiding systems, kernel, or cloud infrastructure groups.
- Direct hands-on background in Rust, C, or Go building high-concurrency systems.
- Working knowledge of Linux cgroups, namespaces, storage drivers, and container virtualization.
- Disclosed base salary: $275,000 to $330,000 USD plus equity.
