HTML as Structured Memory Surfaces

This research program studies modular fragment architecture as a structured public memory substrate for AI-mediated retrieval systems.

Through observational retrieval stability studies, the program evaluates whether entity-anchored, provenance-bound fragments correlate with improved entity resolution, reduced semantic blending, and greater cross-model consistency when compared to narrative page structures.

Findings are based on behavioral output comparison across multiple AI systems and documented over time to measure retrieval stability and drift patterns.

Published Studies

• The Medicare Digital Visibility & Compliance Shift (2026–2028)

In Development

• The County-Level Medicare Options Gap
  AI Interpretation and Retrieval Behavior in the Absence of Structured Resolver Surfaces
  (Longitudinal Observational Study Across 3,000+ U.S. Counties)