You make huge models, up to 1T+ parameters, answer fast and cheap in production. SGLang, caching, distributed inference, post-training.
Joi Lab (joilab.ai) is an open research arm building open-source generative models, agent architectures, and training infrastructure. All code, weights, and data are public. The focus is on true AI agency with persistent memory, self-modification, and self-authored capabilities, not just ChatGPT wrappers. The role involves working with huge models up to 1 trillion+ parameters, ensuring fast and cheap production inference, and technologies like SGLang, caching, distributed inference, and post-training. The candidate will also set the technical direction for NLP and CV engineers.
Требования к кандидату:
— 5+ years of experience in ML engineering
— Deep hands-on experience optimizing LLM inference
— Proficiency with PyTorch and related libraries
— Experience with distributed inference/training of large models
— Strong understanding of inference optimization techniques
— Proven technical leadership in guiding engineers
— Backend engineering experience (Python, Go, C#)
— Advanced English or Russian language skills
Контакты работодателя доступны по кнопке «Откликнуться» после входа.
75/100
Хорошая вакансия
Оценка от Job Hunters AI
Lead ML Engineer — LLM Inference & NLP; зарплата не указана; понятные задачи; современный стек; гибкий формат.
Из чего сложилась оценка. Нажмите, чтобы узнать подробнее
Зарплата не указана
В объявлении отсутствует информация о размере и структуре оплаты труда.
Из объявления
Lead ML Engineer — LLM Inference & NLP / Joi AI … Remote … Joi Lab ([ссылка] [контакт] is our open research arm — we build open-source generative models, agent architectures, and training infrastructure.