PNNL is seeking to develop next generation artificial intelligence solutions for mass spectrometry based omics and real time measurements, specifically focusing on AI foundation models for interpreting raw mass spectrometry signals, multi-agent scientific systems, and machine-learning models for real-time execution on edge devices.
Requirements
- Experience developing algorithms or tools for raw mass spectrometry data (LC-MS, MS/MS)
- Proficiency in Python and experience developing AI/ML data analysis software using ML-related packages libraries (Pandas, Numpy, SciPy, SciKit) and PyTorch for building and training deep learning models
- Experience preparing data for model development, including signal processing or feature extraction for high-dimensional scientific data
- Familiarity with high-performance computing environments and distributed training workflows
- Experience implementing or optimizing ML models on edge hardware such as microcontrollers (e.g., ARM Cortex, ESP32, STM32) or single-board devices
- Understanding of embedded ML techniques such as model quantization, pruning, or conversion to formats suitable for edge deployment (e.g., TensorFlow Lite, ONNX)
- Experience with version control (e.g., Git) and collaborative software development practices
Responsibilities
- Develop algorithms and computational tools for interpreting raw mass spectrometry data across proteomics, metabolomics, and related omics.
- Implement machine learning and deep learning models in Python and PyTorch, including model training, evaluation, and optimization.
- Contribute to the design, training, and deployment of AI foundation models for MS signal interpretation.
- Collaborate with team members to develop multi-agent systems that use foundation models for semi-autonomous, context-aware scientific workflows.
- Prototype and implement ML models for real-time inference on microcontrollers and other embedded or edge-computing platforms.
- Conduct performance optimization for embedded ML, including latency reduction and memory-efficient model architectures.
- Utilize high-performance computing (HPC) systems to train and scale deep learning models.
Other
- BS/BA or higher
- Ability to work effectively in interdisciplinary team environments
- Clear written and verbal communication skills
- Strong problem-solving skills and attention to detail
- Interest in foundation models, multi-agent systems, or autonomous science frameworks
