Software Engineer, Storage Systems (Agent Infra)

present of AI; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Job summary

Baseten is hiring a Software Engineer for Storage Systems (Agent Infra) to help build the distributed filesystem that powers AI sandboxes. The role covers metadata replication, volume placement, garbage collection, and client behavior, requiring deep Linux storage internals knowledge and experience with distributed storage or caching.

As AI agents spawn many sandboxes sharing small files, the engineer must ensure strict isolation, consistency, and low tail latency under heavy multi-mount access.

Qualifications

• Experience with Linux filesystem internals.
• Experience with distributed storage or caching.
• Ability to chase correctness bugs under concurrency/partial failures.

Responsibilities

• Own problems end to end and ship quickly.
• Build the distributed filesystem that agent sandboxes mount (metadata replication, placement, GC, client behavior).
• Address correctness, isolation, and performance under concurrency and partial failure scenarios.

Skills

Distributed storage systems

Go

Rust

C/C++

Linux internals

Tools

Linux

FUSE

Spark

Job description

Software Engineer, Storage Systems (Agent Infra)

This role builds the distributed filesystem that agent sandboxes mount, covering metadata replication, volume placement, garbage collection, and client behavior. A strong candidate has deep Linux storage internals knowledge, hands on distributed storage or caching work, and the ability to chase correctness bugs that appear only under concurrency or partial failure. The team is small, so the person must own problems end to end and ship quickly.

How it fits the AI era

AI agents spin up many sandboxes that share small files at once, so storage must handle heavy multi mount access with strict isolation and consistency. The tooling around models changes fast, but the hard parts here stay human: cache coherency, write ordering, crash consistency, and tail latency under load. AI can speed up routine coding, yet the judgment on correctness and failure modes remains with the engineer.

How it builds AI experience

Working here gives direct exposure to storage built for AI agent workloads, including sandboxes and a filesystem mounted over FUSE. A candidate can build this by studying Linux filesystem internals, trying distributed filesystems or object stores, and contributing to open source storage projects. Profiling and debugging concurrency issues in real systems would strengthen the fit.

Worth showing

• Distributed storage systems
• Go Rust or C/C++

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Baseten

High-performance cloud inference platform for deploying, serving, and scaling open-source and custom artificial intelligence models.

Tools mentioned: Linux, Rust, C++, Spark, machine learning