Engineering Manager, Runtime - Storage
- Compensation
- $165K–$330K / yr
- Location
- San Francisco
- Work setup
- Hybrid
- Seniority
- Not specified
- Employment
- Not specified
- Experience
- Not specified
- Category
- Software engineering
- Posted
Skills
Must have
- Experience managing a team of engineers in a systems, infrastructure, or storage context, including hiring and performance management
- Hands-on background in distributed storage systems, content-addressable storage, or large-scale caching infrastructure
- Strong systems programming background in Go, Rust, and/or C/C++, with the ability to engage credibly in code reviews and architectural discussions
- Working knowledge of the Linux storage stack: filesystems, the page cache, block devices, and mounts
- Understanding of how container images are structured, stored, and delivered at scale
- Strong written and verbal communication skills, with the ability to influence without authority across teams
Nice to have
- Contributions to containerd/containerd, snapshotter projects (stargz, soci, nydus, overlaybd), distribution/distribution, or closely related open-source projects
- Familiarity with lazy-loading snapshotters or peer-to-peer image distribution
- Experience with EROFS, FUSE, ublk, or io_uring
- Experience building or operating CDN, object storage, or large-scale caching infrastructure
- Experience running storage services on Kubernetes
Full description
ABOUT BASETEN
Baseten powers mission-critical inference for the world's most dynamic AI companies, like Cursor, Notion, OpenEvidence, Abridge, Clay, Gamma, and Writer. By uniting applied AI research, flexible infrastructure, and seamless developer tooling, we enable companies operating at the frontier of AI to bring cutting-edge models into production. We're growing quickly and recently raised our $1.5B Series F, led by Altimeter Capital, Conviction Partners, and Spark Capital. Join us and help build the platform engineers turn to ship AI products.
THE ROLE
Container runtimes were designed for general-purpose software workloads. AI inference is not a general-purpose workload.
Running large models at production scale exposes cracks in every layer of the container stack: runtimes unaware of GPU memory constraints, images that take minutes to pull when a model needs to scale to thousands of replicas, and isolation mechanisms that weren’t designed for the multi-tenant serving environments that production AI requires. The tools the industry has relied on for a decade weren’t built for this, and patching around those limitations at higher layers only goes so far.
Baseten owns the entire pipeline, from the moment a developer pushes a model to the moment a request gets a response. That vertical ownership means we can fix these problems at the root. The Runtime Fabric team is doing exactly that: purpose-building the container runtime and storage layers for AI inference workloads, led by some of the world’s top containerd maintainers.
As Engineering Manager for Storage Products on the Runtime Fabric team, you will lead the engineers who build and run the storage products behind that stack: the Baseten Delivery Network (BDN), which sits in the cold start path of every model on the platform, the snapshotters and filesystems that expose model weights and container images to a running container, and the storage that model deployment and training need next. The team is advancing the next generation of caching for inference workloads, from the origin tier to the GPU host. You will own reliability for the area as a measured outcome, report to the head of Runtime Fabric, and stay close to the code through design and code review. If you’ve managed systems engineers and have hands-on experience with caching, content-addressable storage, or distributed filesystems, we’d love to talk.
EXAMPLE INITIATIVES
- How the Baseten Delivery Network (BDN) makes cold starts fast
- Extending BDN from model weights to container image layers, training checkpoints, and deployment artifacts
- Lazy-loading snapshotters and block-level delivery (EROFS, FUSE, ublk) so a container can start before the full image or model has arrived
RESPONSIBILITIES
Team Leadership and Culture