Postdoctoral Research Associate, AI for Operational Safety

Requisition Id 15400 

­­Overview:    

We are seeking a Postdoctoral Research Associate – AI for Operational Safety who will focus on developing AI solutions for several problems related to operational safety at ORNL and DOE sites. This position resides in the Performance Engineering group in the Data and AI Systems Section in Computer Science and Mathematics division within  Computing and Computational Sciences directorate, at Oak Ridge National Laboratory (ORNL).  

 

This role lies at the exciting intersection of Operations and Artificial Intelligence (AI) at scale, with a particular focus on Generative AI for Risk Forecasting, Large Language Models (LLMs), and Human-in-the-Loop AI Systems. Our aim is to advance AI for Operations by integrating next-generation AI agents and LLMs with real-world operational safety and risk management needs at scale. 

Key research areas include AI-driven Risk Analysis, where the focus is on the research and development of large-scale language models and agentic systems for forecasting hazardous events, identifying high-risk infrastructure properties, and supporting human decision-making through natural language explanations. This includes techniques such as instruction-tuned LLMs, reinforcement learning with human feedback (RLHF), and agent-based orchestration for analyzing unstructured operational data. The position further encompasses several problems related to AI for Operations, where the goal is to build scalable, semi-automated, human-in-the-loop AI solutions for DOE and national infrastructure. Focus areas include hazard identification, predictive risk analysis, incident trend forecasting, and dynamic safety modeling using Generative AI techniques. The role will involve working with LLMs and AI agents to process, generate, and reason over large volumes of historical and real-time data. The role also involves Responsible AI, which explores safety-critical LLM alignment, AI robustness under uncertainty, and the development of trustworthy, explainable, and privacy-aware models. 

 

Major Duties/Responsibilities: 

• Collaborate with internal and external researchers on a variety of AI/ML-related research and development projects under the AI for Operations umbrella project. 

• Develop, fine-tune, and evaluate large-scale Generative AI models and agentic systems to support semi-automated, human-in-the-loop systems for AI for Operations. 

• Present and report research results and publish scientific findings in top-tier AI/ML/NLP peer-reviewed conferences and journals in a timely fashion. 
• Deliver ORNL’s mission by aligning behaviors, priorities, and interactions with our core values of Impact, Integrity, Teamwork, Safety, and Service. Promote equal opportunity by fostering a respectful workplace – in how we treat one another, work together, and measure success.

 

Basic Qualifications:

• Ph.D. in Computer Science, Electrical/Computer Engineering, AI/ML, or a closely related field. 

• Demonstrated experience in AI/ML model development, LLM tuning, generative AI, functional safety and risk analysis. 

• Proficiency in Python, with experience in PyTorch, TensorFlow, and common ML libraries (e.g., scikit-learn, pandas). 

• Publication record in peer-reviewed journals or top-tier AI/NLP/ML conferences. 

• Strong communication and scientific writing skills. 

• Ability to work both independently and in large-scale, collaborative research teams. 

• Experience with deep neural networks. 

• Familiarity with NLP packages, pipelines, toolkits (spaCy, NLTK), LLMs and Hugging Face Transformers. 

• Familiarity with Vector databases, RAG and agentic AI framewor