Opensource Geological Tools: Large Language Model AI Assistant for Mineral Resource Analysis and Visualisation

Opensource tool released: Run a Large Language Model driven AI solution locally on your machine for insights into mineral production, reserves, trends, prices, substitutes, and recycling resources.

Mohanty et al (2025) have released, Opensource, the code and data from their research in Github. Link in the comments. It is designed to run with Llama-3-8B a lightweight LLM which can be downloaded from Huggingface to run locally on your machine for privacy. You can add your own documents into the framework. A potentially interesting piece of research to stimulate ideas in a number of areas.

Abstract
The sustainability of advanced technologies depends on the consistent availability of key raw materials. To better understand the availability and risks associated with these materials, we have developed an offline platform that runs locally on a machine, providing AI-driven data analytics and visualization without requiring an Internet connection. Utilizing tools such as Retrieval-Augmented Generation (RAG) with Large Language Models (LLMs) and incorporating geographic visualizations, the platform offers a detailed analysis of material availability and market concentration.

It integrates data from the United States Geological Survey (USGS)
mineral commodity summaries, delivering comprehensive insights into production, reserves, trends, prices, substitutes, and recycling resources. This study focuses on critical materials for energy storage and thermoelectric applications, including lithium, cobalt, nickel, bismuth, tellurium, and rare earth elements, which are essential for renewable energy systems and modern technologies.

The platform enables researchers and policymakers to visualize trends and assess potential risks, supporting informed decision-making and strategic planning to address material scarcity and supply chain vulnerabilities.

Online: https://mineral-ai.net/
GitHub: https://github.com/truptimohanty/Mineral_Viz_RAG
Paper: Mohanty T, Sayeed HM, Mohanty C, Sparks TD. Comprehensive Insights into Global Mineral Commodities: Analysis, Visualization and Intelligent Assistance. ChemRxiv. 2024; doi:10.26434/chemrxiv-2024-vg302-v2