Machine-Assisted Perception for Project Risk Management
"Whoever sees faster, wins." — John Boyd
AI Project Risk Intelligence, Governed by Design
Deterministic Bayesian evidence. A five-layer LLM interpretation hierarchy. Human authority at every decision.
PRIMMS® — the Project Risk Identification Measurement and Mitigation System — pairs a deterministic MATLAB signal layer with an LLM of your choice. The math decides when to speak. The language model decides what to say. The project manager decides what to do.
See How PRIMMS Works
Nine slides from the full presentation — the problem, the architecture, and the governance case.
Hybrid Cognition: The Working Paper
The full theoretical framework behind PRIMMS — Popper, Friston, Bayesian evidence architecture, the Gödel-Turing constraint, and the Zero constraint, operationalized through PRIMMS as a concrete system exemplar.
By John Aaron, PhD. Formalizes why the machine orients and the human decides — not as a design preference, but as a structural requirement grounded in falsificationism, hierarchical predictive processing, and the limits of self-referential inductive systems.
Bring PRIMMS to Your Next Project
See PRIMMS analyse a live project, or run your first real analysis with your own project data.