Inside a Bohaus Sprint: A Medical Billing Office

The situation

The client is an owner-led medical and dental billing company. The owner was a biller first and still works claims. A team of billers and certified coders, a client list of independent physician and dental practices, and a mix of practice management systems on the client side: Tebra, AdvancedMD, Dentrix, Eaglesoft and others. The office works through clearinghouses and payer portals.

Denials arrive every day, through remittance files, clearinghouse reports and portal messages. Each biller works their own clients' denials, usually oldest first or in whatever order they show up. Some denials are small and easy. Some are large and close to the payer's appeal or timely filing deadline. In the queue, they look the same.

The owner sees the result at month end, when a client asks why a large claim was written off.

What the sprint finds

In Discover, we pull a sample of recent denials across clients and sit with billers as they work them. We list every source of denial information and every system a biller opens to resolve one.

The finding: the team is skilled, but the queue has no order. Denial reason codes are not grouped, so the same fixable problem, such as a missing modifier or an expired authorization, gets solved one claim at a time. Deadlines are tracked in each biller's head. Nobody can say what was recovered last month, or what was lost.

In Diagnose, a denial queue ranked by dollars and deadline ranks first. Before anything changes, we set the measures: dollars recovered from denials, denials written off, days from denial to resubmission, and denials that pass their deadline. The aim is to see whether the change pays off, not just to ship it.

What it builds

In Build, we create one denial queue fed by the clearinghouse and remittance data the office already receives.

Each denial comes in with client, payer, amount, reason code and the appeal or filing deadline. AI groups denials by reason and suggests the likely fix and the resubmission or appeal path, drawing on the office's own history and payer rules. The queue ranks by dollars at stake and days left, so large, urgent denials rise to the top.

Coders and billers keep every judgment. They decide what is correct, how it is coded, what is appealed and what is written off. AI drafts appeal letters for a person to review. Protected health information stays in the office's existing compliant systems and clearinghouse.

Denial reasons that repeat for a client are flagged, so the owner can raise them with the practice before they happen again.

Hand over

In Hand over, a senior biller and a coder learn the queue in depth, including the ranking rules and how to override a suggested fix. The rest of the team gets a short session. A playbook covers edge cases, such as a payer that changes its rules mid-year, and names who owns the system.

The owner gets a dashboard of the measures set at the start, broken out by client and payer, plus a 30/60/90-day roadmap with a review date. If the numbers do not move, the review is where that gets said plainly.

What changes

Billers start each day with the denials that matter most. Deadlines stop slipping by. Repeated problems get fixed at the source instead of claim by claim. The owner can show each client what was recovered, and can use that clarity when talking to new practices.

What this means for your business

Set the measures before you change the work. Then you will know whether it paid off.

For a billing office, that is dollars recovered. For a fishing charter, seats filled. For a veterinary clinic, appointments booked from reminders.

That is a Bohaus Sprint: six weeks, fixed scope, four phases (Discover, Diagnose, Build, Hand over), and one live system inside the tools you already use, with a trained team, a playbook, a dashboard and a 30/60/90-day roadmap.