Назад
6 дней назад

Staff Software Engineer (Compute Efficiency)

Формат работы
hybrid
Тип работы
fulltime
Грейд
lead
Страна
UK
vacancy_detail.hirify_telegram_tooltipВакансия из Telegram канала -

Мэтч & Сопровод

Покажет вашу совместимость и напишет письмо

Описание вакансии

TL;DR
Staff Software Engineer (Compute Efficiency) (AI/ML Infrastructure): Design, deploy, and scale observability systems and telemetry pipelines for compute efficiency, hardware health, and workload goodput across distributed clusters with an accent on accelerator utilization, cloud infrastructure, and ML run reliability. Focus on identifying compute waste, integrating new hardware, hardening Kubernetes-based infrastructure, and improving efficiency metrics for training and inference.

Staff Software Engineer (Compute Efficiency)

Company

Isomorphic Labs

Conditions

6 days ago

Skills

Candidate Availability

Required and preferred rules are kept separate and reflect the wording in the original posting.

About the Role

You will design, deploy, and scale observability systems and telemetry pipelines to monitor compute efficiency, hardware health, and workload goodput across distributed clusters. You will find compute waste, improve accelerator utilization and ML run reliability, integrate new hardware, harden cloud infrastructure, and contribute to core technical decisions with ML, platform, and science teams.

Requirements

  • Real-world experience operating, monitoring, and debugging infrastructure for large-scale AI/ML workloads
  • Experience in cloud compute infrastructure design, preferably GCP
  • Strong programming skills
  • Significant experience deploying in Kubernetes at scale
  • Familiarity with Nvidia GPU generations
  • Proven track record building production observability and telemetry stacks

Responsibilities

  • Design, deploy, and scale observability systems and telemetry pipelines for fleetwide compute efficiency, hardware health, and goodput
  • Drive hardware efficiency and node reliability and integrate new hardware
  • Identify compute waste and bottlenecks and optimize accelerator utilization
  • Instrument and improve canonical efficiency metrics for training and inference runs
  • Improve reliability of ML runs
  • Operate, maintain, and harden research, development, and production cloud infrastructure and clusters
  • Collaborate with science, research, product, business development, and operations teams
  • Contribute to core technical decisions on tooling, infrastructure, and architecture

Benefits

  • Hybrid working model

Будьте осторожны

Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →

Текст вакансии взят без изменений

Источник -