REVENUE CYCLE
Claims, denials, payer workflows, and AR.
Working inside the day-to-day mechanics of getting healthcare revenue from “something happened” to “this is actually resolved.”
ABOUT
I learned revenue cycle before I started teaching AI how to do it.
I didn’t come to healthcare AI from machine learning. I came to it from the claim, the denial, the payer portal, the workflow, and the people trying to make all of that move.
PHOTO / TBD
Annie
doing Big Girl
Computer Stuff
doing Big Girl
Computer Stuff
CURRENT LANE
REVENUE CYCLE + AI
THE SHORT VERSION
I’m an RCM operator who ended up in AI because the systems needed someone who knew what the work was supposed to look like.
My background is in healthcare revenue cycle operations, prior authorization, claims workflows, payer processes, and operational leadership.
I’ve spent years working inside the systems, queues, payer rules, documentation requirements, and operational decisions that sit between a healthcare service being delivered and the revenue actually making it where it needs to go.
Along the way, I became increasingly interested in the systems behind the work: why workflows break, where specialists rely on judgment instead of a checklist, what can be standardized, and what absolutely cannot.
Eventually, that curiosity collided with AI.
BEFORE THE AI
First came the operations.
The AI work makes a lot more sense once you understand where the domain knowledge came from.
PRIOR AUTHORIZATION
Clinical requirements meet payer reality.
Navigating authorization rules, payer-specific requirements, documentation, deadlines, and the operational consequences when any of those pieces go sideways.
OPERATIONS + LEADERSHIP
People, process, volume, and scale.
Leading teams and building workflows that still have to function when the queue gets ugly, the volume grows, the edge cases show up, and real people have to use the process every day.
PROCESS
Looking past the task to the system around it.
Understanding where work slows down, where knowledge lives only in someone’s head, where handoffs fail, and where a better process can remove friction without removing judgment.
THE SHIFT
Then the question changed.
It stopped being only, “How should a specialist do this work?”
It became: how do you teach a system to understand enough of the work to do it well?
That opened up a completely different set of problems.
If an AI agent is investigating a claim, what information does it need before it can make a decision? How should it behave when the evidence is incomplete? What does a correct result actually look like? How do you know when something sounds reasonable but is operationally wrong?
Those questions pulled me directly into AI enablement, evaluation, agent workflows, and QA.
WHERE I AM NOW
Somewhere between the RCM specialist, the AI system, and the people building it.
Today, my work sits at the intersection of revenue cycle expertise and AI systems designed to support or perform that work.
I translate domain knowledge into something product, engineering, and AI systems can use. I test what comes back. I identify the gaps between what the system did and what an experienced specialist would have done. And I help turn those gaps into better workflows, stronger evaluation, and more reliable behavior.
I’m especially interested in the messy middle: the place where a system can technically work and still not be good enough for the real operational environment it’s being asked to enter.
That’s the part I like.
CREDENTIALS
CRCR
Certified Revenue Cycle Representative
CPC
In progress
DOMAIN AREAS