Ever since I graduated college, I wanted a job that paid me multiple six figures. I did everything: worked in finance, took the courses, moved into tech, and eventually broke through. Then I broke through again. Multiple multiple six figures, on paper a big number, and in some sense it really is.

But here's the weird part. My life didn't really change. Not between $200K and $300K, and honestly not much past that either.

What did change, slowly, was the work itself. The problems started feeling personal, and I was putting in twelve-hour days without wanting to stop, which was ironic because I'd assumed more money would push me toward more work-life balance and the opposite happened. I kept asking myself why. Was it the prestige ladder I'd climbed before turning 27? Some kind of superiority complex? Eventually it hit me. I'd stumbled into something that felt like purpose, and purpose is a hungrier thing than money.

So naturally, I quit.

You're probably thinking this guy's lost it. He just told me he loved his job and now he's leaving? Maybe a little. But before you close this email, let me tell you what I figured out, because it's the thing that wouldn't let me stay.

Healthcare providers in the United States now spend roughly $60 billion a year on revenue cycle management, the unglamorous machinery of billing, coding, prior auth, and claims, and that number is climbing about ten percent a year. On top of that, providers lose another eight to ten percent of their revenue to denials and write-offs, which means doctors and hospitals are paying tens of billions to get paid, and still leaving tens of billions more on the table.

Which is the question that wouldn't leave me alone.

I'll come back to that. First let me show you the story that taught me where to look.

A patient, let's call him David, walks into Dr. Singh's office with a stomach ache. Dr. Singh runs the usual checks and decides David's blood should go to a lab called Nyra for a genetic test that screens for cancer markers. Nothing exotic, just the kind of test that's become more common as cancer rates have climbed.

David has insurance through a big-name payer we'll call BHC. Before the test can happen, BHC requires something called prior authorization: special permission, granted in advance, confirming the test is what they call medically necessary. (We'll spend a whole future issue on what that phrase actually means, because spoiler, it doesn't mean what you think.)

Nyra has a partnership with Dr. Singh's office, so they handle the prior auth on his behalf. They collect David's clinical notes, pull together everything BHC says it wants, and send the packet over. Then they wait. Ten days later BHC says yes. David gives his blood, the test runs, and it catches an early tumor marker so Dr. Singh can start treatment before things get worse.

That is the best case scenario. Read it again. Ten days of delay, a patient sitting at home stressed, a billing team that spent hours assembling a packet for a test everyone clinically agreed was the right call from minute one.

And here's where the question I asked you earlier starts getting an answer.

For a long time I assumed denials happened for the reasons the denial letters said: missing documentation, lacks medical necessity, wrong code, the standard biller mistakes that better training would supposedly fix. That's the explanation the industry tells itself, and it's the explanation most RCM vendors will sell you a tool to fix. Then I started actually looking at the denials, and I noticed something weird.

The biller hadn't gotten lazy. The biller had submitted exactly what BHC's policy required last quarter. The policy had been updated three weeks earlier and the new version asked for a different clinical note, in a different format, justifying a different criterion, and nobody told the lab or the doctor. The denial said "missing documentation," but that wasn't really what happened. What happened was that BHC had quietly moved the goalposts and then penalized everyone who didn't notice.

Once I saw that pattern, I started seeing it everywhere. Denials marked "lacks medical necessity" where the policy text hadn't changed but the medical reviewer was scoring it differently this month. Tests that didn't need PA last year and suddenly did. The denial reasons looked like biller errors, but they were actually policy moves the biller had no reasonable way to track.

That's when the model snapped into place.

Per CMS, doctors spend about thirteen hours a week and $34,000 a year just dealing with prior auth, and the AMA reports PA delays care for the majority of patients who go through it. Now you know why. None of this is a bug, it's the design. Each of these three moves is small on its own, but stacked across thousands of policies and dozens of payers, they form a wall that providers can't keep up with on their own.

Which brings us back to the question I planted earlier.

I spent months trying to answer that one, and the honest answer turned out to be more frustrating than the question

I talked to over a hundred RCM vendors before starting my company, and most of them don't really care about the underlying problem. They sell on the promise of collecting more revenue, then service the contract by hiring teams overseas to do the work as cheaply as possible. The bigger ones have slick marketing about AI and data science, and to be fair some of those tools genuinely work, but most of them break constantly because none of these companies are building the foundation underneath: an actual, current, queryable understanding of payer policy.

And payers know it. They know the people on the other side of the wall don't have a system that can keep up with policy changes, so the wall keeps moving, the denials keep landing, and providers keep paying their eight to ten percent to vendors who can't actually fix what's broken. The vendors aren't really the villains here, they're just running the playbook the industry built, and that playbook is missing the foundation.

That's the gap I left to fill.

The company is called Converus AI, and we're building the policy intelligence layer that should have existed all along. Every clinical policy, every admin policy, every PA requirement, in one place, current, queryable, and self-updating, with AI-native RCM services bolted on top, starting with prior auth. We're focused first on genetic testing, oncology, and endocrinology because that's the space I know cold.

pre·imbursed is the public side of all of it. Think of me as your friend on the inside of the industry, the one who's seen how payer policies actually work and is willing to walk you through it in plain English. Whether you're a patient holding a denial letter and wondering whether to fight it, the CFO of a lab trying to understand why your collections are slipping, a VP of reimbursement at a health system trying to make sense of why your team is drowning, a billing manager, a founder, a caregiver, an employee staring at an EOB you can't decode, you're in the right place. The system is the same system for all of you. The vocabulary changes, but the moves don't.

If a system that quietly moves the goalposts and then penalizes everyone who didn't notice sounds like something worth understanding, stick around.

That's issue one. Hit reply — I read everything, and the best questions become the next lesson.

— Shivang