Probabilistic Hazard Information (PHI) and AI/ML Research Specialist (Subject Matter Expert II)

Overview

Rapidly updating, probabilistic severe weather guidance is only as good as the science and technology behind it. As FWI's PHI and AI/ML Research Specialist supporting NSSL, you will apply expert, highly specialized knowledge to advance Probabilistic Hazard Information (PHI) systems by integrating Warn-on-Forecast (WoFS) ensemble guidance, meteorological observations, and machine learning techniques into next-generation short-term hazard prediction tools. Your work will bridge research and operations — from designing AI/ML methodologies to leading Hazardous Weather Testbed (HWT) engagement with real NWS forecasters — to make severe weather guidance faster, smarter, and easier to act on, while providing technical guidance and limited supervision to project team members. Work Schedule and Location: Remote: This full-time remote position will work Monday through Friday, 8 AM CST to 5 PM CST.

Responsibilities: * Apply expert, specialized knowledge to advance PHI research and development activities that improve rapidly updating probabilistic forecasts of severe weather hazards * Design, develop, and evaluate AI/ML methodologies that enhance the generation, interpretation, and communication of probabilistic hazard forecasts * Lead development and implementation of tools that integrate WoFS ensemble output with PHI systems and workflows, providing technical guidance to team members * Develop methods to merge WoFS forecasts with meteorological observations (radar, satellite, lightning, environmental datasets) to create enhanced short-term forecast guidance products * Conduct data analysis, visualization, verification, and explainability studies to evaluate forecast performance and improve user understanding of AI-assisted forecast guidance * Direct development of software tools, workflows, and visualization products to support research, testing, and operational transition activities * Lead planning, facilitation, and analysis of user engag