Research Engineer, Fundamental AI Research (fair) - Generative Models/llm Acceleration

Meta Meta · Big Tech · Tel Aviv, Israel

Research Engineer focused on advancing generative models and LLM performance through pioneering algorithmic research, large-scale training, optimization, and evaluation. The role involves leading initiatives, collaborating across teams, and contributing to publications and open-source projects.

What you'd actually do

  1. Innovate, lead, and execute pioneering algorithmic research to push the state-of-the-art in generative models and LLM performance
  2. Directly contribute to the experimental process, including designing details, implementing reusable code, running evaluations, and organizing results
  3. Collaborate with cross-functional teams (research, product, infra) to build new and advance LLM optimization
  4. Analyze and optimize code for quality, efficiency, and performance, and provide feedback to peers during code reviews
  5. Lead initiatives, provide technical guidance and mentorship to peers, and help onboard new team members

Skills

Required

  • Master's degree or higher in a relevant technical field (e.g., Computer Science, Machine Learning, AI, or related discipline)
  • 6+ years of experience in machine learning, deep learning, or AI research, or equivalent practical experience
  • Experience designing and implementing large-scale model training pipelines using frameworks such as PyTorch or JAX
  • Experience with distributed computing and parallel training techniques including data parallelism, model parallelism, or pipeline parallelism
  • Experience debugging and optimizing AI systems for performance, reliability, and correctness across the full model lifecycle
  • Experience building evaluation frameworks and benchmarking pipelines to measure model quality and capability regressions
  • Experience with large language model pretraining, fine-tuning, post-training, or inference optimization

Nice to have

  • Contributions to peer-reviewed AI research publications or open-source AI frameworks

What the JD emphasized

  • 6+ years of experience in machine learning, deep learning, or AI research
  • Experience designing and implementing large-scale model training pipelines
  • Experience with distributed computing and parallel training techniques
  • Experience building evaluation frameworks and benchmarking pipelines
  • Experience with large language model pretraining, fine-tuning, post-training, or inference optimization
  • Track record of contributions to peer-reviewed AI research publications or open-source AI frameworks

Other signals

  • push the state-of-the-art in generative models
  • LLM performance
  • large-scale model training pipelines
  • distributed computing
  • large language model pretraining, fine-tuning, post-training, or inference optimization