aipythonkubernetesmachine learningmodel evaluationmodel fine tuninggpu infrastructurellm inference
Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
ЗП не указанаразвернуть
На похожих ролях по рынку
$182к$209кмедиана$254к
48 в выборке
На СНГ рынке: $4.6к/мес ($4.1к - $4.7к)
90
Крутая вакансия
развернуть
У роли четкий senior-уровень, серьезная ответственность за новую AI-платформу и работа с востребованной инфраструктурой моделей; основные компромиссы - офисный формат и контекст проп-трейдинга.
Работа из офиса
Современный AI-стек
Высокая ответственность
Востребованная сфера
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Senior AI (Infrastructure) Engineer (Open-Weight Models/LLM Infrastructure): Building the model layer for open-weight model training, self-hosted inference, evaluation, and intelligent model routing with an accent on production gateways, GPU infrastructure, and model quality, latency, and cost controls. Focus on designing distillation pipelines, fine-tuning and serving large language models, and operating reliable Kubernetes-based inference systems.
Location: Office-led working in the United States, including New York, Illinois, and Chicago
Company
hirify.global represents a global proprietary trading firm building a centralized AI capability and reducing dependence on external frontier-model providers.
What you will do
Build and operate a production model gateway routing inference across open-weight and commercial models with fallback, rate controls, and quality, latency, and cost visibility.
Design distillation pipelines and training datasets for smaller task-specific open models.
Fine-tune and productionize open-weight models such as Llama, Qwen, and Mistral for internal workloads.
Deploy and maintain self-hosted inference with vLLM, TGI, Triton, or comparable technologies, focusing on reliability and throughput.
Build and optimize GPU infrastructure on Kubernetes, including provisioning, scheduling, monitoring, workload management, and troubleshooting.
Develop model evaluation and regression frameworks and define standards for model selection, governance, and future agentic workloads.
Requirements
6–10 years of software, machine learning, or AI engineering experience, including strong production-level Python and ownership of technically demanding systems rather than primarily integrating third-party APIs.
Meaningful production experience fine-tuning, post-training, or distilling open-weight models, including knowledge of training data, methodology, evaluation, failure modes, and production trade-offs.
Hands-on experience serving large language models in self-managed infrastructure with vLLM, TGI, Triton, or equivalent inference frameworks.
Strong experience managing GPU-backed production environments, including provisioning, scheduling, capacity planning, utilization monitoring, performance optimization, and operational troubleshooting.
Deep familiarity with Kubernetes for GPU-intensive machine learning workloads and the infrastructure requirements of reliable model serving.
Experience building model evaluation, benchmarking, and regression-testing approaches covering quality, latency, cost, and production behavior.
Nice to have
Experience with quantization, PEFT, LoRA, or similar techniques for reducing training and inference costs while preserving production quality.
Exposure to model gateways, inference proxies, Hugging Face, the open-model ecosystem, or AI systems in financial services or other sensitive-data environments.
Culture & Benefits
Permanent position in a small dedicated AI team with significant influence over architecture, standards, and technical direction.
Greenfield, highly technical, well-capitalized environment within a proprietary trading firm.
Office-led working arrangement in the United States.
Hiring process
Apply through the role application form; all fields marked with an asterisk are required.
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 →