Enterprise Architecture & AI Systems Research
Official research and technical specifications created by Danh Nolan. Establishing deterministic systems across both scales: macroscopic website architecture with Decision Ladder Navigation™ (DLN™) and microscopic LLM prompt control with the RCCOF Framework™.
Structures entire websites around the 5 sequential decision rungs of human buyers (OCPRA™) and 4 disciplined link rails. Eliminates dead-ends, keyword cannibalization, and funnel friction.
Explore DLN Specification5-layer deterministic prompt control (Role, Context, Constraint, Output format, Few-shot) for enterprise AI content systems. Eliminates statistical hallucinations and enforces brand voice.
Explore RCCOF SpecificationFramework 01: Information Architecture
Traditional websites organize navigation by departmental silos or abstract sales funnels, resulting in dead-ends, keyword overlap, and frustrated users. DLN assigns every URL exactly one decision role and every page exactly one logical forward step.
| Evaluation Factor | Legacy Marketing Funnels | Decision Ladder Navigation™ (DLN™) |
|---|---|---|
| Page Purpose | Pages try to do multiple jobs at once (explaining, proving, and selling on one URL). | One decision role per URL. If a page needs two rungs, it is split into two distinct pages. |
| Next Step | Dead ends, generic "Read More" links, or aggressive popups on introductory content. | One clear next step per page leading directly to the next logical decision rung. |
| Search Intent | Multiple blog posts and service pages compete against each other for the same query. | Distinct rungs prevent internal keyword overlap and clarify topical hierarchy. |
| Price Transparency | Hiding pricing behind "Contact Us" forms, leaving high-intent visitors frustrated. | Dedicated Rate rung explaining cost drivers, pricing models, and scope boundaries. |
Framework 02: Prompt Engineering & AEO
Large Language Models operate as probabilistic token predictors. When prompts lack structured constraints, models regress to the statistical median, generating generic, cliché-ridden, and hallucinated prose. RCCOF enforces a 5-layer control architecture to guarantee enterprise-grade fidelity.
| Evaluation Dimension | Unstructured / Ad-Hoc Prompting | RCCOF Framework Specification |
|---|---|---|
| Tone & Style | Fluctuates wildly; repetitive clichés and conversational filler. | Deterministic and disciplined; calibrated to exact Brand Voice via Few-shot. |
| Factual Integrity | Frequent confabulation and plausible-sounding hallucinations. | Strictly bounded by Context ground truth and negative Constraint fences. |
| Structure & AEO | Monolithic walls of text requiring heavy human editing. | Modular, snippet-ready sections engineered for Google AI Overviews and citation. |
| Team Reusability | Individual prompt crafting; results cannot be reproduced across team members. | Centralized prompt library where any writer produces identical caliber outputs. |
The Dual Engine Architecture
High-performing enterprise digital systems require coordination across both scales: the macroscopic navigation map and the microscopic generative precision.
DLN determines what job a URL must accomplish, who the visitor is when they land, and which specific next step they must take. It eliminates multi-purpose confusion across the website topology.
RCCOF translates the page's DLN role into strict generation parameters: injecting the domain role, ground-truth context, negative constraints, answer-first formatting, and approved few-shot exemplars.
Technical Publications
Access the complete open documentation suite for both frameworks:
The core technical definition covering the five rungs, the four rails, cluster tiers, metric formulas, and Rasmussen disambiguation.
Read documentationThe official open technical specification for deterministic prompt engineering: Role, Context, Constraint, Output Format, and Few-shot.
Read documentationHow to assign domain-specific identity and authority to large language models to control response depth, tone, and eliminate hallucinations.
Read documentationHow the Orient rung resolves conceptual confusion without premature sales pressure, forming the foundation of content clusters.
Read documentationTen evaluation criteria comparing DLN™ against TOFU-MOFU-BOFU and See-Think-Do-Care, documenting failure modes and selection boundaries.
Read documentationProduction metrics from danhnolan.com documenting architectural changes, indexing coverage, and variables that were intentionally left out.
Read documentationFive cost variables that drive restructuring work, with a unit-count estimate that transfers between teams without guessing hours.
Read documentationThe practical rollout sequence with un-gated CSV downloads: the DLN Labelling Sheet and the 5-Column Link Matrix.
Read documentationLicensing & Rights
Decision Ladder Navigation™, DLN™, OCPRA™, and the RCCOF Framework™ are trademarks of Danh Nolan. All original specifications, tables, and diagrams are Copyright © 2026 Danh Nolan. All rights reserved.
Free commercial application: You are free to apply DLN™ and the RCCOF Framework™ in your internal enterprise architectures, commercial products, and client consulting work without fee or prior licensing authorization.
Academic citation: When referencing, researching, or building upon these frameworks, please cite the official Zenodo deposits:
Nolan, D. (2026). Decision Ladder Navigation (DLN): An Intent-Driven Information Architecture Framework. Zenodo. https://doi.org/10.5281/zenodo.22843346Nolan, D. (2026). RCCOF Framework: A Structured Prompt Engineering Specification for Enterprise Content Generation. Zenodo. https://doi.org/10.5281/zenodo.22856974Author registry: Danh Nolan · ORCID: 0009-0007-7906-5091.