We are looking for a Staff Software Engineer to join our Unified Compute team and take technical ownership of the low-level systems that power how scientific workloads run in the cloud. Our scheduling project (Sched) has gained real momentum, and we need a senior engineer to help build the ecosystem around it to the next level. The core of this role is deep in the stack — Fusion (a FUSE filesystem presenting cloud object storage as POSIX), Fusion Snapshots (CRIU-based checkpoint/restore that survives spot-instance reclamation), and Ignite (a container process manager) — with Sched as the orchestration layer that ties compute together. You'll also support adjacent services such as Groundswell (ML-based resource prediction) and Cloudinfo (multi-cloud pricing/metadata). It's a hands-on senior role with strong architectural influence
Design, build, and own low-level systems across Fusion, Fusion Snapshots, Ignite, and Sched
Write reliable, well-tested code primarily in Go and C, with Rust and JVM (Groovy/Java, Micronaut) where the codebase calls for it
Work close to the OS: FUSE filesystems, Linux namespaces, CRIU checkpoint/restore, process supervision, signal handling, static musl builds
Tackle multi-cloud orchestration and scheduling: provider abstractions (AWS, Azure, GCP), instance selection, autoscaling, spot-instance fault tolerance, durable async queues
Drive performance engineering — chunked I/O, multi-level caching, buffer pools, bandwidth management
Set technical direction and raise the standard for testing, observability, and operational excellence
Mentor teammates and help onboard new engineers