Software Engineer, Compute and Isolation

present of AI; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Baseten is seeking a Software Engineer, Compute and Isolation to own the full lifecycle of a microVM used by AI agents. You will manage boot, snapshot, fork, and resume, and optimize performance across diverse host hardware.

We require hands-on experience with KVM and hypervisors such as Firecracker or QEMU, deep Linux internals, and a focus on multi-tenant isolation. Strong debugging from userspace to kernel and hardware is essential.

Qualifications

• Hands-on KVM or hypervisor work and Linux internals expertise.
• Experience with multi-tenant isolation and security across software layers.
• Strong debugging across userspace, kernel, and hardware.

Responsibilities

• Own the full lifecycle of a microVM (boot, snapshot, fork, resume).
• Tune performance on diverse host hardware and optimize isolation.
• Debug across userspace to kernel to hardware.

Skills

Kernel debugging

Linux internals

Performance tuning

Multi-tenant isolation

Systems programming

Tools

KVM

Firecracker

QEMU

Linux

Rust

Go

Job description

Software Engineer, Compute and Isolation

This role asks for someone who can own the full life of a microVM that agent sandboxes run on, from boot and snapshot to fork and resume, and tune how it performs on each host's hardware. A strong candidate has hands on work with KVM and a hypervisor such as Firecracker, Cloud Hypervisor, or QEMU, plus deep Linux internals knowledge across namespaces, cgroups, memory, and the boot path. The person should have built multi tenant compute as the provider, treating isolation as a security boundary, and be comfortable debugging from userspace through the kernel to hardware.

How it fits the AI era

AI agents now write and run code, call tools, and hold state across long tasks, so the machines under them must start, pause, snapshot, and resume quickly while keeping tenants apart. Much of the tuning and debugging still depends on human judgment about kernels, hardware, and isolation tradeoffs. The role sits close to the infrastructure that makes agent workloads practical at scale.

How it builds AI experience

Working here means gaining direct experience running the compute layer that AI agents depend on, including fast VM resume, snapshot and fork, and multi tenant isolation on bare metal. That knowledge is built by shipping and debugging real virtualization systems rather than by studying models. Contributions to open source hypervisor and kernel projects would strengthen a case for this work.

Worth showing

• KVM and Firecracker
• Multi tenant isolation
• Rust or Go systems
• VM snapshot and resume

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