I'm Ken Mendoza, an independent researcher and scientific software engineer working where information theory, geometry, dynamical systems, and the messier parts of the living world overlap. One idea runs through all of it: most complex systems hide a shorter description than their surface behavior suggests, and a surprising amount of useful work comes from finding it.

I build computational methods rooted in information theory and Riemannian geometry, then test them on real data through artifact-backed case studies, across formal mathematics, critical-transition detection, biomedical signals, and materials. The posture is deliberately in-silico and evidence-bound: public datasets, reproducible runners with checksums, and explicit claim ceilings. When a hypothesis is falsified, it's reported openly rather than quietly dropped.

Before this came decades across computational science and software: startup technical leadership (I co-founded Digital Lava, which went public in 2000), biomedical and proteomics R&D at Arbor Vita, where I was the sole computational scientist and am a named co-inventor on three issued United States patents, and a long run as an independent systems architect. I studied microbiology and political science at UCLA, and began graduate study in government at Cornell, with the philosophy of science in my first semester.

I am the author of The Republic in the Machine: Why AI Agents Need a Last Day in Office, narrative nonfiction on why delegated power, human or machine, needs an ending by default. It is a complete working draft under author review; the proposal and sample chapters are available to literary agents on request. The research program behind it lives at MendozaLab.org.

The lab-grade evidence (formal-conjectures contributions with their stated proof scope, the Erdős Atlas workbench, manuscripts and datasets) lives at Mendoza Laboratory. This is the short version; the full CV has the detail.

Full curriculum vitae: choose a format