Trust, but verify, every payer call

Davner is a managed prior authorization service. We handle the work from submission through determination, including calling payers on behalf of the practice when the authorization requires it.

Representing a practice in those conversations requires trust. It shouldn't require blind trust.

Previously, practices could see our summary of each payer call. That was enough to understand the result, but they still had to take our word for exactly what was said. A summary is useful precisely because it collapses information, and sometimes the information it leaves out becomes important later.

Users can now open the complete transcript for every payer call we make. They can check our work, trace the exact language used in a confusing response, or review the conversation if there's ever a dispute with the payer.

The answer at a glance

Most payer calls are boring. A call might spend an hour repeating that all representatives are assisting other callers and giving the same estimated wait time again and again.

We've used plenty of annoying medical software that has no concept of what information a person can actually use. We don't want someone opening an authorization and getting hit with walls and walls of text when the real question is simply, “What happened?”

Every call still gets a brief summary in the case timeline. It includes the information staff generally need in their day-to-day work: who was reached, the result of the call, the authorization or reference number, and anything that needs to happen next. The timeline also shows the number called, when the call occurred, and how long it lasted.

That's enough for most cases. The complete information is available through a “View transcript” link instead of competing with the useful summary.

The full conversation when it matters

The transcript presents the call as a dialogue, with Davner and the payer clearly labeled. Staff can read through the exact conversation instead of trying to reconstruct it from a short note.

That matters when the payer's wording is ambiguous, when a staff member wants to double-check the AI's work, or when a dispute arises later. Payers already preserve their side of these conversations. Giving the practice access to the complete transcript helps even the playing field.

It also creates a level of operational visibility that practices rarely have when calls are made manually. A person generally writes a note at the end of the call. Whatever didn't make it into that note is gone from the practice's record. AI calls produce the full transcript without asking anyone to do extra work, so there's no reason for us to collapse that information permanently.

Executives can audit how Davner represents their organization to payers. Authorization staff can inspect an individual answer when they need to move a case forward. Most people won't open most transcripts, and that's the point: transparency should be available without turning into information overload.

A second review before moving forward

The transcript also supports another safeguard inside Davner.

One AI system prepares the call summary. A second, independent agent reads the complete transcript, the summary, and the context of the authorization before the result moves the case forward. Its job is to identify potential conflicts, such as a summary, status, authorization number, or payer statement that doesn't match the conversation.

If something needs another look, it can be escalated to a person on our team. Davner is a managed service, so the answer to uncertainty isn't to quietly pass it on to the practice.

Every practice would like to have multiple people double-checking every authorization. Most don't have enough time in the day to do it. One advantage of AI is that we can cross-check this work without making staff listen to calls or read transcripts that don't require their attention.

The promise is simple: trust, but verify. Davner represents practices to payers, and practices can see exactly how we're doing it.

Complete payer call transcripts are live now in the Davner case timeline.