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
$103к$144кмедиана$184к
21 в выборке
На СНГ рынке: $4.8к/мес ($3.9к - $8к)
93
Крутая вакансия
развернуть
Роль дает редкую возможность влиять на обучение передовых моделей, их оценку и внедрение в продуктовой компании; задачи подробно описаны и сосредоточены на AI.
Передовые методы
Понятные задачи
Сильное направление
Открытые исследования
Поддержка переезда
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Senior Machine Learning Engineer (AI): Designing and operating large-scale training and post-training systems for frontier language models, agents, and production APIs with an accent on distributed training, evaluation, and model serving. Focus on diagnosing unstable training and regressions, building synthetic-data and LLM-as-judge pipelines, and improving latency, throughput, memory use, and serving cost.
Location: Copenhagen, Denmark
Company
hirify.global builds AI models, agents, APIs, and infrastructure for critical industries, with a focus on healthcare and developer-controlled systems.
What you will do
Design and run distributed training systems covering architecture, data pipelines, parallelism, checkpointing, evaluation, and recovery.
Improve model capabilities through supervised fine-tuning, reinforcement learning, and distillation.
Develop and evaluate models for agentic work, including tool use, evidence handling, persistent context, and reliable task completion.
Create and calibrate evaluations, including LLM-as-judge systems, to expose failure modes and validate improvements outside the training distribution.
Turn stronger models into practical APIs by improving latency, throughput, memory use, and serving cost.
Requirements
Experience training large language models and owning technically complex machine learning work.
Strong practical distributed-training experience and depth in areas such as mixture-of-experts models, reinforcement learning, or distillation.
Hands-on experience with post-training methods such as DPO and GRPO, including diagnosing reward hacking, unstable runs, and noisy signals.
Experience building training-data pipelines, ideally synthetic-data pipelines, and understanding how training data changes model capabilities.
Production experience evaluating LLMs and building and calibrating LLM-as-judge systems.
Ability to investigate new models, papers, and agent frameworks through experiments and comparative testing.
Ability to write maintainable code for other engineers and explain technical trade-offs clearly.
A relevant advanced degree is welcome, but equivalent experience is equally valued.
Nice to have
Experience serving open-weight models.
Experience with agents building tools, coordinating with other agents, or investigating agent failures.
Culture & Benefits
Full-time position reporting to the CTO.
Equipment provided.
Relocation support provided.
Research is published, benchmark resources are released, and platform methods are documented for engineering inspection.
Different backgrounds and paths into machine learning are welcomed; substantial relevant experience can compensate for not matching every listed item.
Hiring process
Submit a link to something built, a difficult problem solved, or a result that deserves to be challenged.
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 →