Назад
Company hidden
2 дня назад

Lead/Senior Engineer, Model Inference (AI)

Формат работы
hybrid
Тип работы
fulltime
Грейд
senior/lead
Английский
b2
Страна
UK/Switzerland
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Lead/Senior Engineer, Model Inference (AI): Building and operating Mariner, a production model-serving platform for Thomson Reuters' custom and benchmarked large language models with an accent on inference optimization, reliability, observability, and infrastructure sovereignty. Focus on designing scalable distributed serving systems, reducing cost and latency through batching, caching, and quantization, and establishing compliance, capacity-planning, and incident-management practices.

Location: Hybrid in Zug, Switzerland, or London, United Kingdom

Company

hirify.global Labs is the company's research and development division, building state-of-the-art AI products and custom models for regulated professional workflows.

What you will do

  • Build and operate Mariner, the model-serving infrastructure powering production traffic for the Thomson model family and benchmark models.
  • Implement and tune inference optimizations such as batching, caching, and quantization to reduce cost and latency without sacrificing quality.
  • Improve observability, reliability, disaster recovery, and operational performance as call volume and deployment coverage grow.
  • Implement ISO 27001 controls and evidence, cost tracking and attribution, and GPU-capacity contingency plans.
  • Work with training and evaluation teams to deploy new model releases and document cost, latency, and reliability tradeoffs.
  • Establish on-call, incident-review, onboarding, and other engineering practices as an early member of the team.

Requirements

  • Hands-on infrastructure engineering experience with Kubernetes or a similar platform, including production ownership.
  • Experience operating model services across peak load, scaling, disaster recovery, and operational-excellence requirements.
  • Direct experience with model-serving infrastructure and/or inference optimization for large models.
  • Large-scale distributed-systems experience, including scaling and reliability under production load.
  • Ability to debug distributed systems and production incidents under real traffic.
  • Track record of designing, building, and operating production software end to end.
  • Ability to champion and enforce engineering best practices while responding to dynamic product requirements.
  • Ability to operate autonomously as an early lead engineer and mentor and grow team members.

Nice to have

  • Experience with vLLM, TensorRT-LLM, Triton, SGLang, or similar LLM-serving stacks.
  • Exposure to GPU-cluster capacity planning or multi-region and multi-cloud deployments.
  • Experience in regulated or compliance-heavy production environments such as ISO 27001 or SOC 2.
  • Familiarity with systems-development languages such as Rust or Go.
  • Experience with systems-level optimization, rate limiting, routing, and bottleneck analysis across networking, hardware, and lower-level software.

Culture & Benefits

  • Flexible hybrid working environment for office-based roles.
  • Work-from-anywhere flexibility for up to 8 weeks per year.
  • Flexible vacation, two company-wide mental health days, retirement savings, tuition reimbursement, incentive programs, and wellbeing resources.
  • Learning and professional development programs focused on continuous skills growth.
  • Two paid volunteer days annually and opportunities for pro-bono consulting and ESG initiatives.
  • Values-driven culture focused on customers, rapid learning, inclusion, work-life balance, and collaboration.

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 →