Sitema ARKON

ARKON is a materials-science engine for analyzing products made from many kinds of materials, identifying performance gaps, and developing improved formulas, mixtures, and processes. It evaluates conventional materials alongside industrial, semi-industrial, and household waste residues as potential inputs, checking them against the properties each product requires. ARKON combines scientific-neuron calculations; product, formula, and rule data in Supabase; normalized material-property data in Google Sheets; and a Turso library of technical evidence extracted from engineering and science books, manuals, scientific studies, tests, and product data sheets. Meta Llama 3.2 is the central AI brain: property-specific neurons tell it what material, state, property, and evidence to investigate; Llama searches Turso and returns relevant source-backed findings to those neurons. ARKON’s neurons then continue the analysis using the system’s scientific formulas and rules. In this demo, ARKON analyzes a glass-fiber-reinforced cement board, flags seven potential issues, distinguishes missing properties from recorded values, and displays scientific cross-formulas and possible improvement paths. I am a self-taught founder who completed primary school; Meta AI tools helped me build ARKON’s software and document-extraction workflow.