I'm the sole architect of KAIA (Knowledge Architecture for Intelligent Agents), an independent research project exploring whether meaning can be represented geometrically rather than statistically, and whether that foundation can make AI accessible without expensive hardware.

Since April 2026, I've completed 27 experiments validating the architecture across a seven-benchmark suite: 85% agent routing, 897,000 memory queries/second, and 100% geometric convexity, all held across 15 consecutive experiments, on standard CPU hardware with no GPU required.

My work has independently converged with findings from cognitive science (Gärdenfors), neuroscience (hippocampal geometry, 2026), and LLM interpretability (Zhou et al. 2025), three fields that reached the same geometric conclusion from completely different directions.

I publish all findings openly, including the honest limitations. The research is designed to be reproducible on any hardware.