TL;DR
Deep Learning Software Engineer (AI): Designs, develops, and optimizes AI software and frameworks with an accent on distributed algorithms, computational graph transformation, and performance bottleneck identification. Focus on optimizing code for various computing hardware backends and interacting with machine learning researchers.
Location: Onsite in Shanghai, PRC
Company
hirify.global is a technology corporation focused on driving revenue growth and market expansion by leveraging its product portfolio, empowering its sales force, and ensuring strategies are targeted and impactful.
What you will do
- Design, develop, and optimize AI software and frameworks.
- Contribute to external AI frameworks such as PyTorch and SGLang.
- Implement various distributed algorithms for machine learning and deep learning frameworks.
- Transform computational graph representations of neural network models.
- Develop machine learning and deep learning primitives in mathematical libraries.
- Profile distributed deep learning models to identify performance bottlenecks and propose solutions.
Requirements
- Master's or Ph.D. degree in Computer Science, Artificial hirify.globalligence, Software Engineering, or related fields.
- Good programming skills in modern C++ and Python.
- Knowledge of Deep Learning theory foundations and some hands-on experience.
- Fluent English communication (both written and spoken).
- Passion for solving problems and a positive mindset.
Nice to have
- Experience in performance optimization or high-efficiency kernel development.
- Experience with PyTorch, SGLang, or vLLM.
- Experience with LLM and deep understanding of model structure.
Culture & Benefits
- Committed to Responsible Business Alliance (RBA) compliance and ethical hiring practices.
- Provides ample opportunities for career advancement and skill development.
- No recruitment fees are charged during the hiring process.
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 →