AI Framework Software Engineer

Intel Intel · Semiconductors · Shanghai, China

AI Framework Software Engineer at Intel focused on designing, developing, and optimizing AI software and frameworks. The role involves implementing distributed algorithms, transforming computational graphs, developing ML/DL primitives, profiling models, and optimizing code for hardware backends. Collaboration with ML/DL researchers is key.

What you'd actually do

  1. Conducts design and development to build and optimize AI software.
  2. Designs, develops, and optimizes for AI frameworks (e.g., SGLang) and contribute to public upstream.
  3. Implements various distributed algorithms such as model/data parallel frameworks, parameter servers, dataflow based asynchronous data communication in machine learning, and/or deep learning frameworks.
  4. Transforms computational graph representation of neural network model, and develops machine learning and/or deep learning primitives in mathematical libraries.
  5. Profiles distributed deep learning models to identify performance bottlenecks and proposes solutions across individual component teams.

Skills

Required

  • Master degree or Ph.D. in Computer Science, Artificial Intelligence, Software Engineering, or related fields
  • Modern C++ and Python programming skills
  • Foundations of Deep Learning theory
  • Fluent English communication (written and spoken)

Nice to have

  • Experience of performance optimization or high efficiency kernel development
  • Experience of PyTorch, SGLang, vLLM
  • Experience of LLM and deep understanding of model structure

Other signals

  • optimize AI software
  • AI frameworks
  • distributed algorithms
  • performance bottlenecks
  • machine learning frameworks