Breaking 4 Years of Stagnation at CCAoA
What Actually Happened
CCAoA had spent 4 years and multiple consultants trying to modernize military childcare verification. Nothing shipped. Military families waited 88 days on average with 76% having to resubmit - not because they made errors, but because nobody told them what to upload in the first place. This triggered months-long email loops. They hired me for my documentation expertise to figure out why nothing worked.
The Ask
Document how the system actually worked. That was my mandate. They brought me in because I had a reputation for documentation that could untangle messes. They wanted to understand why a project started and tabled years earlier kept failing.
Here's what I found: "Simple" applications required 40-60 manual fields scattered across broken systems, Excel spreadsheets, sticky notes, and processes that only lived in operations' heads. Complex families? 80-120 fields across all these disconnected places. The goal got adjusted to just get ONE family through by March, then make it great by October.
But my documentation revealed the real problem: 76% of these "requirements" weren't even contractual. They were layers of workarounds and patches that nobody designed but everyone was forced to live with - 20 years of accumulated complexity.
"The feedback I got when I wrote this contract was that it was 100% more detailed than we've ever written any scope for a contractor." - COO/CTO
What I Did
My documentation became the blueprint for transformation.
I created a three-lens analysis:
Contract requirements: What DoD actually required
System reality: 70% of code wasn't even used. Quarterly income multiplied by 24 instead of 4. Fundamentally broken.
Operational practice: Operations doing "2, 3, sometimes 4 touches" per application because the data lived in their heads and spreadsheets
This revealed the truth: most complexity wasn't required. It evolved through 20 years of patches on patches.
When they wouldn't give me research resources, I flew myself to Charleston and watched a military family try to navigate this mess. That single session showed why 76% resubmit. Families don't know which version of their LES to upload. They can't find providers that supposedly exist. They get lost bouncing between systems, spreadsheets, and phone calls. This direct observation unlocked everything.
Once I understood what actually mattered versus organizational overhead, I could fix it:
Cut application complexity by 75%:
Simple families: 40-60 scattered fields → 15-20 smart questions
Complex families: 80-120 fields → 30-40 guided interactions
Military sponsor section alone: 22 steps → 7 conditional screens
The breakthrough wasn't removing requirements. It was hiding complexity through intelligence:
"Branch → Component → Status" replaced 22 manual fields across three spreadsheets
"Does your spouse work?" skips 10 irrelevant fields
"Upload your LES" extracts 15 data points automatically instead of manual entry
Plain language for spouses, military terms for service members
Instead of just getting one family through, I delivered a complete working application by March 2025 - 7 months before the October deadline. Built 5-10 prototypes daily. Showed stakeholders their tribal knowledge turned into working software. While they evaluated Tootris for millions, I built competing prototypes on weekends.
Created data sync to legacy systems so we could migrate gradually. Each release adds intelligence while keeping current operations running. No $5M big bang risk. Just continuous improvement that brings staff along instead of threatening them.
The Results
Delivered working software in 3 months after 4 years of nothing:
Exceeded the "get one family through" goal - delivered complete application by March 2025
9 successful releases, zero critical bugs
Set foundation to reduce 76% → 25% resubmission rate
Processing time path: 88 days → eventual 15 minutes
Transformed how government serves people:
"No wrong answer" design - families get guided, never blocked
OCR/AI extracts data from documents automatically
Dynamic document lists show exactly what's needed, when
Conditional logic prevents errors before they happen
Built sustainable change:
Operations went from "you never listen" to "Judson's been a dream"
Verification specialists will verify, not re-enter data from spreadsheets
Each release makes their expertise more valuable, not obsolete
The real breakthrough? Documentation isn't about understanding problems. It's about seeing solutions others miss. I documented reality so thoroughly that denial became impossible. Then we built what military families actually needed, not what 20 years of accumulated patches demanded.
Currently expanding the March success with enhanced flows for Q3. The organization still debates vendor contracts while we're shipping real solutions - one intelligent question at a time.
What's next?
Tools that amplify human expertise instead of replacing it. OCR extracts data, families confirm it, specialists apply judgment where it matters. When a military spouse faces deployment at 2am, that's when human help becomes critical. The system gets smarter about when to automate and when to connect people.
No $5M transformation. No consultant armies. Just rapid prototyping and incremental wins that prove the model. Each release makes operations more valuable while serving families better.
