AI Scientist

GE Healthcare GE Healthcare · Healthcare · Bellevue, WA +2 · Digital Technology / IT

AI Scientist at GE Healthcare focusing on generative AI, large-scale pretraining, prompt tuning, distillation, robustness, responsible AI, and quantization for healthcare data. The role involves developing and implementing novel ML algorithms for LLMs to automate clinical tasks using EMRs, waveforms, and clinical reports, demonstrating algorithms, exploring human feedback, building prototypes, and working with large-scale datasets. Requires a Master's or Ph.D. with relevant research experience, publications in LLM/Agentic AI/SSL, and expertise in building large-scale AI models.

What you'd actually do

  1. Developing and implementing novel machine learning algorithms particularly in the area of LLM to provide automation of clinical tasks using one or more of electronic medical records, waveforms, and clinical reports.
  2. Demonstrating algorithms to meet accuracy requirements on general subject population through statistical analyses and error estimation.
  3. Exploring learning from human feedback and assisting humans evaluating AI.
  4. Building prototypes to enable development of high-performance AI algorithms in scalable, product-ready code.
  5. Working with large-scale datasets, designing, and developing generative algorithms.

Skills

Required

  • Master's Degree in a “STEM” major or equivalent field plus 3 years of relevant research OR Ph.D. in a “STEM” major or equivalent field with 3 years of relevant research.
  • Publications as first author on LLM, Agentic AI or self supervised learning (SSL).
  • Demonstrated expertise in building large scale AI such as generative AI models.
  • Implementation experience with a variety of high-level languages (e.g. Python, C++)
  • Experience with high-dimensional imaging data and waveform/time-series data.

Nice to have

  • Experience and demonstrated capability to handle challenges with vague or abstract problem definition.
  • Experience with frameworks and tools such as DeepSpeed, HuggingFace, Megatron, PyTorch lightning, etc.
  • Experience with various MLOps, ModelOps, FMOps (Foundation Model Ops) methods.
  • Experience working with large scale AI training.
  • An in-depth understanding of machine learning algorithms and modeling (e.g., semi-supervised or weakly supervised learning, generative models, transfer learning, optimization, large language models, etc.)
  • Track record in developing machine learning solutions using massive real-world data for solving real world business problems.
  • In depth experience with Spark/Hadoop and either PyTorch/Tensorflow
  • Experience creating production environment data analytics and applications

What the JD emphasized

  • Publications as first author on LLM, Agentic AI or self supervised learning (SSL).
  • Demonstrated expertise in building large scale AI such as generative AI models.
  • Experience with high-dimensional imaging data and waveform/time-series data.

Other signals

  • Generative AI
  • LLM
  • Pretraining
  • Prompt Tuning
  • Distillation
  • Quantization
  • Responsible AI