Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
развернуть
$140к-200к/год
$160к$200кмедиана$250к
По похожим ролям (46) ~ $200к/год
91
Крутая вакансия
развернуть
Чётко очерченная роль в мультимодальном ИИ сочетает передовые исследования с внедрением в продакшен и даёт сильный технический опыт в быстрорастущей сфере.
Передовой домен
Современный стек
Полный цикл работы
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Member of Technical Staff, Multimodal Foundation Models (Robotics/Embodied AI): Building vision-language and video models and production pipelines that analyze egocentric and manipulation footage, annotate video, and detect quality issues with an accent on multimodal model training, post-training, and distributed GPU experimentation. Focus on designing scalable temporal-grounding and captioning systems, validating model outputs against human audits, and turning research results into reliable production systems.
Location: Hybrid in New York, Toronto GTA, or San Francisco
Salary: Toronto: CA$140K–CA$200K; New York and San Francisco: $140K–$200K per year
Company
Build data infrastructure for robotics and embodied AI, including data capture, labeling, and hardware-enabled workflows used to train and validate humanoid and embodied AI systems.
What you will do
Develop and improve vision-language and video models for egocentric and manipulation footage.
Build model pipelines that identify defects, missing coverage, and task-correctness issues in captured footage, validating results against human audits.
Post-train models using supervised fine-tuning, preference optimization, and reinforcement learning to improve reliability and reasoning.
Design scalable pipelines for video captioning, temporal grounding, and annotation.
Run large experiments on distributed GPU infrastructure and convert results into production systems.
Requirements
Hands-on experience training, fine-tuning, or post-training large vision-language or multimodal models.
Strong deep learning fundamentals and experience with PyTorch or JAX.
Experience designing, running, and analyzing large experiments.
Familiarity with distributed training and GPU-based development.
Strong research instincts and engineering skills, including building clean systems and taking experiments into production.
Nice to have
Experience using vision-language and video models for footage quality control, defect detection, and identifying gaps in captured data.
Experience with structured captioning, temporal grounding, and scene understanding.
Experience with reward modeling, synthetic data, and data curation and filtering for multimodal models.
Culture & Benefits
Full-time hybrid work in New York, Toronto GTA, or San Francisco.
Inclusive work environment welcoming applicants from all backgrounds and committed to equal opportunity.
Reasonable accommodations are available during the recruitment process.
AI-assisted tools may support application evaluation, while human judgment remains part of the review.
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 →