Member of Technical Staff - Inference Runtime

Modal is building the infrastructure layer that AI workloads need to run at scale. The company powers category-leading organizations like Ramp, Cognition, DoorDash, and Suno with instant GPU access, sub-second container starts, and native storage capabilities. Modal recently raised a $750M Series D at a $15.75B valuation and has grown 15x in annualized run rate over the past year. The engineering team includes open-source creators, academic researchers, and experienced infrastructure leaders who are rethinking how AI applications deploy and run in production.

The Inference Runtime team is seeking a systems engineer to own the container runtime stack that powers inference and training workloads across Modal's fleet. You'll work at the intersection of Linux, containers, filesystems, storage, GPU drivers, and distributed systems. This role sits directly on the critical path for inference performance—improvements here can reduce cold starts, increase token throughput, and unlock new accelerator types. You'll shape the architecture of Modal's container platform and contribute to open-source runtime technology used by every inference and training workload.

What you'll do

• Dramatically accelerate container startup, checkpoint, and restore times for large inference and training workloads by designing zero-copy and low-copy data paths between container memory, filesystems, storage, and runtime components.

• Build multi-GPU and accelerator-aware snapshotting systems that efficiently handle GPU memory and RDMA-enabled workloads, with optimization techniques including parallel uploads, direct I/O, and incremental snapshots.

• Improve container image and filesystem performance across EROFS, FUSE, page caches, overlay filesystems, and remote storage systems.

• Extend the sandboxed runtime to support new GPUs, drivers, profiling tools, and device capabilities across NVIDIA and AMD hardware.

• Debug complex failures across system calls, virtual memory, process lifecycle, kernel behavior, GPU drivers, and container isolation.

• Safely roll out runtime and kernel changes across a heterogeneous fleet using compatibility controls, feature flags, and observability tools.

• Work across runtime, scheduler, storage, and GPU infrastructure teams to investigate production problems and deploy solutions.

What you'll bring

• Strong Linux systems knowledge including processes, virtual memory, filesystems, system calls, scheduling, namespaces, cgroups, and signals.

• Experience building or debugging container runtimes, sandboxes, low-level infrastructure, or the Linux kernel.

• Strong programming ability in Rust, Go, or another systems language, with willingness to become productive in both Rust and Go.

• Track record profiling and optimizing systems where memory movement, I/O, synchronization, or kernel interactions are performance bottlenecks.

• Ability to debug across abstraction boundaries from application behavior through runtimes, drivers, and kernel.

• Skill turning loosely defined production problems into well-designed, reliable systems.

• Desire to own foundational infrastructure and collaborate closely with internal teams and customers.

Nice to have

• Experience with gVisor, runsc, runc, OCI runtimes, seccomp, or checkpoint/restore systems.

• Linux kernel development, virtualization, sandboxing, kernel modules, or device proxying.

• CUDA, ROCm, GPU drivers, accelerator virtualization, or GPU profiling.

• Contributions to open-source systems software or experience with large-memory and multi-GPU inference systems.