Seek hidden connections

Cue Cloned Notes Audit Tool
A screenshot from Cue's "Cloned Notes" visit note audit tool
(Cue is Ambiki's predecessor that was built for a specific practice)

(Post 4 of 5) This is the fourth in a series of posts about the origins of Ambiki’s values.

Our fourth value: “Seek hidden connections”.


One unfortunate aspect of the therapy industry is insurance fraud. While it's a minimal occurrence, it does happen. A classic example is when a therapist bills for a session that never took place.

For this reason, it's essential for practice owners to implement self-audit and compliance policies and procedures. One such procedure I've seen involves the audit team manually checking visit notes for "cloned notes", a common red flag for fraud. Fraudulent therapists might lazily copy and paste the same note repeatedly.

However, this method immediately made me think, "There has to be a better, more automated way."

During a walk, I pondered industries where document similarity is crucial. Then, a lightbulb moment! In my previous job at TM-Town, we assessed document similarity to match specialized translators with suitable projects.

That realization led to an idea: why not use natural language processing to detect similarities in visit notes? By flagging notes with high similarity, we could streamline the audit team's process. Instead of sorting through all visit notes or randomly sampling, they could focus on the flagged ones.

Cue visit note audit tool
The audit team can easily compare flagged notes side by side

And thus, our "cloned notes" tool was conceived.

At Ambiki, we believe the best ideas typically come from unexpected associations when broadening our field of vision.