FURI | Fall 2025

Agentic AI for Automated Data Extraction of Energy Materials and Devices

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

This research project investigates the extraction, organization, and comparison of data from published energy materials studies using agentic artificial intelligence. In order to evaluate the accuracy of AI-derived and original datasets, the project uses large language models like Gemini, Claude, ChatGPT, and Copilot to automate data extraction from scientific literature. In order to produce structured data on interfacial and operational characteristics among classes of energy materials, the method uses reasoning-based extraction techniques. Reducing manual data processing and facilitating quicker, more accurate model development for upcoming energy research are the goals of this work.