Staff Software Engineer, GPU Infrastructure Lifecycle Management at Together AI
San Francisco, CA
$240,000–$280,000from the description
Jul 15, 2026
San Francisco, CA
Jul 21, 2026
What this job asks for AI summary
This role centers on building the software that automates the full lifecycle of physical hardware — from bare-metal discovery through GPU stack bring-up, health validation, and decommission — so that inference and research teams can stand up or tear down clusters via a single API call rather than manual tickets. The work involves designing declarative control planes, provisioning state machines, durable workflow orchestration, and self-healing automation. It suits engineers with a background in orchestration systems, event-driven architectures, and a product-oriented approach to internal platforms.
Senior level · National · Full-time
“or” means any one of them counts — you don't need all of them.
Posted 2 times — it's one opening, so apply once.
We read this from the posting text with AI. Skim the description below before ruling yourself out.
How this req sits in the market our data
Rare in this occupation — lead with these, and say what you built with them.
Most people in this occupation already list these. Still required — just not what gets you shortlisted.
What the occupation pays Median $138,970 (middle half $107,524–$175,762). This posting is about at that midpoint.
Estimated from BLS employment for this occupation and area, per-skill prevalence across our listing corpus, and published wage benchmarks — as of Jul 28, 2026. It is a model, not a headcount.
Why we read it this way (7)
No location is specified in the posting; the CBSA and state fields are left blank. The compensation range is stated as a US base salary, confirming a US-based role.
The title is a bare 'Software Engineer' with no level modifier; advertised seniority is Unspecified. However, the scope — owning a full provisioning control plane, designing state machines and declarative APIs, and operating them in production — supports a Senior classification.
The primary language requirement is 'Go, Python, Rust, or similar'; Go is used as the canonical name with Python and Rust as named alternatives. The JD also separately calls out 'Systems programming in Rust or Go' as a Nice to have, which is consistent with Go/Rust being the preferred languages.
Temporal and Cadence are listed as interchangeable examples of durable workflow orchestration tools under the core requirements section.
Kafka, NATS, and SQS are listed as interchangeable examples of event-driven messaging under the core requirements section.
Kubernetes controllers/operators are cited as an example of the required control-plane/orchestration experience; Kubernetes is treated as a hard gate on that capability.
All bare-metal provisioning (PXE/iPXE, Redfish/IPMI, BMC), networking (VLANs, BGP), and GPU stack (NCCL, CUDA, InfiniBand/RoCE) skills appear exclusively under the 'Nice to have' section.
Read the full posting
The employer publishes the full description on their own site — read it there ↗. Or sign in to read it here — it's free, and it also lets you track this application.