Skip to content
Work Tech Weekly
Nataly Youssef

How AI Can Catch the Medical Billing Errors Your Employees Never Will

A woman went to the emergency room for what would be diagnosed as sepsis and cellulitis. Even though she'd gone to a preferred network and done everything her plan asked of her, the intake was coded as “pain in the left leg.” The result? A $100 copay turns into a $4,000 bill. 

The woman was charged as if the visit had never been a true emergency. Catching the mistake wasn't anyone's job. At least, it wasn’t anyone’s job until Nataly Youssef built an entire company focused on addressing medical billing errors like these.

Nataly is the Founder and CEO of Reclaim Health, a platform that uses AI to comb through employer claims and benefits data, surface the money employees are owed, and act on it without waiting for anyone to file a complaint. Nataly didn't start out trying to fix medical billing. She began her career optimizing patient flow inside hospital systems before moving on to advise a claims data warehousing firm. That’s where she watched the billing side of healthcare up close and saw how often the math doesn’t favor the patient. And, for an employee who makes $35,000 a year, the impact of a $4,000 error can be devastating.

We talked about why medical billing errors are a predictable output of a fragmented system rather than a rare glitch, why more benefits communication isn't the fix, and how AI can get an employee's money back once an error occurs.

How AI Can Catch the Medical Billing Errors Your Employees Never Will
  36 min
How AI Can Catch the Medical Billing Errors Your Employees Never Will
Work Tech Weekly
Play

The Medical Billing Error Nobody Catches

Nataly's not shy about where the blame sits. The U.S. healthcare billing system was built for payers, not patients. Employers end up inheriting all that complexity without ever being handed the tools to manage it.

She watched this problem come into focus as healthcare data went from five percent digitized at the start of her career to 95 percent by the time she finished her PhD. “Not even hospitals knew... what negotiated rates they should be negotiating in the first place,” she says. If even hospitals can't answer what something costs, what chance does the average person have of catching an error on their own?

That's the pattern behind the sepsis bill. A true emergency got coded at intake as something minor, and every event followed from the mistake. “That really created for me the urgency to focus on the financial burden of the problem, not just the efficiency problem of it,” Nataly says.

Why Employees Don't Fight Back — And Why That's the Wrong Frame

The standard response to a confusing benefits experience is more communication. Nataly thinks that’s the wrong response. A fatter open enrollment packet isn’t an answer. The woman who got hit with the $4,000 bill understood how to use her benefits. She went to a network facility. 

A coding issues can be buried in details that are barely visible to the trained eye of an expert. It’s unlikely that someone will become fluent in claims adjudication in the middle of a hospital stay. As a result, the typical response when someone who doesn’t understand the hieroglyphics of claims language is to settle.

“What they start looking for is ways to just pay it because they're worried that this would be on their credit,” she says. Paying an incorrect bill has less to do with not knowing better. It’s more about simply not having the time to fight it or fear of any negative consequences.

This is why Nataly draws such a hard line between information and action. An app that surfaces insights still leaves the burden on the employee. “[Reclaim] is not just AI for insights,” she says. “This is AI that is helping take action and that is removing the burden of taking action on employees that are working long hours.”

It's also the reason 9 out of 10 employer groups now choose Reclaim's opt-out model over opt-in. When you make this the default setting, an employer sends a signal: we want this to work without creating more work.

What Reclaim's Claims Intelligence Actually Does

Reclaim's model connects an employer's claims data with its benefits data, then calculates the gap between what an employee owes and what they should owe, automatically. Nataly compares it to what TurboTax did for personal taxes. “It's not like the tax code has been fixed, right?” she says. Nothing about the underlying complexity changed. Reclaim flags the error, gets it fixed with the claims administrator, and puts the money back — then tells you it happened.

It even catches the stuff nobody else would bother chasing. “Not missing a $75 that is owed to someone,” as Nataly puts it, is less about the dollar amount and more about what it signals. If the system only catches large errors, employees quietly absorb the cost of every small one, including the sepsis bill, which got corrected once Reclaim Health flagged the miscoding.

The burden of catching and disputing an error shouldn't belong to ordinary people who work full time, don’t have a map of the medical coding minefield, and don’t want to spend an afternoon on hold with a claims department. For an employer, this is the difference between a benefit that exists on paper and one an employee actually experiences. For employees who may wonder if AI is actually on their side, it’s a strong trust signal.

The sepsis bill got fixed only because Nataly happened to be watching for it. That's the strongest argument in favor of Reclaim's model. The burden of catching an error shouldn't sit with the person least equipped to catch it.

A system that detects billing mistakes automatically and starts the recovery before anyone has to ask isn't just a healthcare fix. It's proof that AI can work for the person it's supposed to serve, not just around them.

Let's Talk