Senior Data Scientist III – Generative AI & Foundation Model Development

Overview

At PNNL, our core capabilities are divided among major departments referred to as Directorates. The AI and Data Analytics Division is part of the National Security Directorate and combines domain expertise with advanced hardware and software to deliver computational solutions that address complex data and analytic challenges. PNNL is a key partner in the Department of Energy Genesis Mission to develop a powerful AI platform to accelerate scientific discovery, drive energy innovation, and strengthen national security.

Read more about AIDA: https://(Use the "Apply for this Job" box below).-analytics

We are seeking a Senior Data Scientist – Generative AI & Foundation Model Development (level
3). This role blends foundational research and real-world implementation to solve cutting‑edge challenges in modern AI. The core technical focus includes development, adaptation, and deployment of large language models (LLMs) and other generative AI systems, involving hands‑on work with model internals, training pipelines, and optimizer‑level tweaks.

As a Senior Data Scientist, you will serve as a technical leader, acting as principal investigator and/or project manager, driving proposal development, and engaging directly with stakeholders. The position is based in either Richland, WA or Seattle, WA and requires onsite presence.

• Conduct cutting‑edge research to develop and refine generative AI techniques, including large language model pretraining, fine‑tuning, agentic systems, and novel training methodologies.
• Take ownership of defined tasks or small to medium projects, proactively identifying technical challenges and proposing solutions to senior staff or project leads.
• Drive research execution by designing, building, and iterating on foundation models and agentic AI systems that address complex real‑world problems.
• Apply deep expertise in model internals (tokenization, attention mechanisms, weight adaptation, and loss optimization) to customize and improve model performance for domain‑specific applications.
• Evaluate AI and agentic systems across multiple axes, including standard performance metrics, uncertainty quantification, generalization to out‑of‑distribution settings, and mission‑specific assurance criteria.
• Serve as a mentor to junior staff or interns, fostering a collaborative and inclusive research environment.
• Engage with stakeholders to understand project requirements and translate them into actionable technical approaches aligned with sponsor goals.
• Support proposal development and business development by contributing to proposals and/or briefing sponsors and building effective working relationships within and across interdisciplinary teams.
• Translate cutting‑edge research papers and findings into mission‑relevant insights, tools, and prototypes for stakeholders and operational users.
• Maintain awareness of emerging trends in AI, AI security, and national security to shape future research directions and identify new opportunities for impact.

• Knowledge of current ML research landscape, especially xAI, adversarial machine learning, AI safety, and the science of deep learning.
• Knowledge of compute environments and their cybersecurity concerns.
• Strong communication skills with the ability to deliver technical findings to diverse audiences.
• Demonstrated proficiency in creating proposals and technical reports.
• Proven ability to collaborate effectively with a multi‑disciplinary team environment.

Minimum Qualifications:

• BS/BA and 5+ years of relevant work experience
• MS/MA and 3+ years of relevant work experience
• PhD with 1+ year of relevant experience

Preferred Qualifications:

• Degree in computer science, statistics, mathematics, or related fields.
• Experience in generative AI, foundation model, agentic system development in applied science/research for national security challenges.
• Hands‑on experience analyzing the internal structures of deep learning models, particularly large language models and large vision models.
• Experience in quantifying uncertainties in neural networks and complex systems.
• Strong experience in Python development and common…