Research
The architecture is not an opinion.
OrbisFramework is built on a research thesis. The thesis is published. The methods are documented. The work continues.
Featured Paper
The Decision Cannot Be Averaged
Why conventional AI scoring methods fail for high-stakes decisions, and how alignment-based and weakest-link architecture addresses the failure modes. The theoretical foundation of OrbisFramework decision gates.
Read the white paperThe Book
The Architecture of Decision
The complete framework for building AI systems that make high-stakes decisions. Seven components, four edges, and the multiplication principle that separates systems that work from systems that fail. Grounded in 24 centuries of decision theory and validated against frontier AI research.
Learn more about the bookResearch Streams
Decision Architecture
Alignment-based and weakest-link methods for high-stakes decisions. Why averaging fails when getting one dimension wrong makes the whole decision wrong.
Innovation Alignment
Dissertation research at Daniels College of Business, University of Denver. How organizations align innovation investments with strategic objectives.
Convergence
How 24 centuries of decision theory and frontier AI research arrive at the same architectural requirements. The seven components were not invented; they were recovered.
All Papers
The Decision Cannot Be Averaged
April 2026Analysis of why conventional weighted-average scoring methods produce systematically incorrect results for high-stakes decisions, with proposed alignment-based alternatives.
Read paperEnterprise AI Infrastructure Patterns
Coming 2026Architectural patterns for enterprise AI workflow systems, with focus on orchestration, audit, and human-in-the-loop decision gates.
Coming soonFor the full academic record, including publications and ongoing research:
Visit bradleywpetersen.com/researchResearch meets practice
See the research applied to real workflows.
A strategic conversation about how the methodology applies to your situation.