Baseten

Software Engineer- GPU Fabric Observability

Baseten
Apply
2 months ago

Base Salary

$200k - $380k/yr

Responsibilities

  • Own Baseten’s GPU fabric observability and root-cause analysis architecture.
  • Build telemetry pipelines across switches, NICs, hosts, GPUs, Kubernetes, and inference services.
  • Model topology, flow paths, service ownership, and failure domains.
  • Separate true fabric faults from host, NIC, GPU, kernel, driver, RDMA, scheduler, and workload failures.
  • Create operator workflows for triage, remediation, and post-incident learning.

Requirements

  • Staff-level or senior staff-level experience building production infrastructure software.
  • Strong distributed systems background, especially in streaming systems, telemetry pipelines, diagnostics, or control-plane software.
  • Experience processing high-volume, high-cardinality, noisy operational data.
  • Understanding of networking fundamentals and high-performance networks.
  • Ability to work with low-level infrastructure signals and build practical correlation, anomaly detection, or root-cause analysis systems.

Benefits

  • Competitive compensation including meaningful equity, with no base salary amount stated.
  • 100% medical, dental, and vision insurance coverage for employees and dependents.
  • Flexible PTO and company-wide Winter Break from Christmas Eve through New Year's Day.
  • Paid parental leave.
  • Fertility and family-building stipend through Carrot.
  • Company-facilitated 401(k).
  • Exposure to a variety of ML startups and related learning and networking opportunities.

Tech Stack

Categories

Baseten

About Baseten

201-500 employees

Inference is everything. Baseten is an AI infrastructure platform giving you the tooling, expertise, and hardware needed to bring great AI products to market - fast. Our proprietary Inference Stack utilizes the cutting-edge of performance research combined with highly performant and reliable infrastructure to give you out-of-the-box global availability with 99.99% of uptime.