Kai
Dunham

Building Fraxia

Critical Minerals are AI's oil. Fraxia tells you where to drill.

Madison, WI

About

About

Hi, I'm Kai Dunham. Industrial & Systems Engineering @ UW-Madison, building Fraxia. Interested in AI orchestration, agentic systems, semiconductors, manufacturing, and complex adaptive systems.

I build systems that transform unstructured disclosures and datasets into structured, probabilistic signals for real-world investment and operational decisions, with memory layers that compound every signal into sharper judgment over time.

Projects

Elkano Dev

AI diligence pipeline built for a live critical minerals PE fund. Extracts structured deal intelligence from NI 43-101 technical reports, reducing analyst data entry from 3 hours to under 5 minutes.

Fraxia

The AI intelligence layer for critical mineral capital allocation. Elkano Dev was the prototype — Fraxia is the company. Turns mining disclosures into probabilistic risk scores for institutional investors globally, with a memory layer that compounds every signal into sharper intelligence over time. Same pipeline that cut analyst data entry from 3 hours to under 5 minutes.

  • Meridian

    Algorithmic prediction market engine actively trading across event-driven markets. Models probabilistic outcomes and executes data-driven strategies in real time.

  • FabIntel

    Semiconductor intelligence system that translates fab-level disclosures into operational signals. Uses Monte Carlo simulation to probabilistically quantify manufacturing execution risk.