PATTERN 158

Finding elegant solutions to complex problems

I cheat, but I cheat fair.

I Reverse-Engineer Chaos Into Clarity

After 28 years building and rescuing enterprise systems, I've learned that the best solutions aren't always the obvious ones—they're the elegant shortcuts that work properly.

Legacy System Rescue

Forensic reverse-engineering of "black box" systems. Making undocumented legacy systems work when documentation doesn't exist.

Cross-Platform Integration

Building bridges between systems that weren't designed to talk. SCORM, xAPI, AICC protocol expertise.

AI Implementation

Thoughtful AI/LLM application. Understanding when to use generative AI vs. structured data for production systems.

Systems Architecture

Building foundational systems that harmonize—like the diapason register that gives a pipe organ its character.

Featured Projects

⚠ FINDING 1

CSBB Dispatch Framework

2011-2016 | Status: 10+ years in production

Analysis

Energy utilities needed cross-domain SCORM delivery before browsers or LMS vendors supported it. Same-origin policy blocked the obvious solutions.

Solution

Engineered cross-domain communication using EasyXDM. Built SCORM API proxy for transparent AICC/SCORM protocol translation. Created automated package generation tooling.

Outcome

System served FPL, Exelon, NRG Energy, BP, Calpine, Colorado Springs Utilities, and Puget Sound Energy. Remained in production 10+ years.

JavaScript EasyXDM SCORM AICC Groovy
⚠ FINDING 2

Ready2Role Legacy Rescue

2004-2016 | Status: Extended system lifespan 10+ years

Analysis

Inherited undocumented courseware CMS on Macromedia JRun platform when original developer left. No documentation, no source code access, production system serving clients.

Solution

Decompiled Java/JSP packages to understand system behavior. Built JavaScript workarounds where server-side refactoring wasn't practical. Maintained critical system beyond expected EOL.

Outcome

System remained operational 10+ years. Enabled company to continue serving clients while planning proper migration path.

Java/JSP JavaScript Macromedia JRun Forensic Engineering
⚠ FINDING 3

Microsoft MCAPS AI Training

2025-2026 | Status: Production

Analysis

Needed AI instructor agent in Copilot Studio. Prompt engineering alone couldn't establish reliable guardrails for progress tracking and navigation.

Solution

Built data pipeline converting Articulate Rise materials into structured JSON. Used AI code interpreter for content transformation, structured data for deterministic control. Created adaptive cards for navigation and assessments.

Outcome

Reliable course navigation and progress tracking. Demonstrated hybrid AI/structured data architecture pattern.

Microsoft Copilot Studio Adaptive Cards AI Code Interpreter JSON

The "Cheat Fair" Philosophy

Like a tuning fork finds the right frequency, I find the optimal solution—but always within the rules.

The "Cheat" Part

  • Using AI to analyze 34,650 emails instead of reading manually
  • Decompiling code when documentation doesn't exist
  • Building cross-domain proxies instead of waiting for vendors
  • Finding the 2-hour solution vs. the 2-week rewrite

The "Fair" Part

  • Validating AI output against source data
  • Testing decompiled insights before production
  • Documenting workarounds for maintainability
  • Ensuring shortcuts aren't technical debt

The Influences

  • Air crash investigation (forensic methodology)
  • Speedrunning/TAS culture (optimal paths)
  • Pipe organs (harmonic foundation)
  • The Three Stooges (direct problem-solving)