Provider organizations are continuing to adopt AI tools across their departments and clinics, and ambient documentation startups are facing the challenge of differentiating themselves in a field that’s become quite crowded with scribing solutions.
The next phase of ambient documentation goes beyond simply scribing — it’s about using visit transcripts and context to do useful work like coding intake and care coordination, according to one executive in the space. During Reuters’ Total Health conference in Chicago, Heidi Health CEO Tom Kelly said the real opportunity lies in transforming ambient AI from a passive listener into an active assistant that actually gets tasks done before and after the visit.
He noted that Heidi is focusing on “solved experiences.” The company aims to deliver specific solutions to defined problems — such as a large volume of unanswered patient intake calls or incomplete post-visit check-ins — rather than build a general-purpose AI platform that tries to do everything, Kelly explained.
“We have some organizations starting to use Heidi for before-visit conversations about scheduling, and also intake. It asks questions like ‘Why are you coming in today?’ or ‘Do you have insurance?’ and says things like ‘Hey, this practice only covers these types of visits’ or ‘These are the fees you’re likely to get,’” he stated.
In addition to automating work, long-term winners will also offer tools that clinicians genuinely enjoy using and rely on daily, Kelly declared.
Heidi’s technology is designed to be simple and require very few clicks, he added. It allows clinicians to assign tasks directly to Heidi with minimal setup, with the goal of driving adoption from the bottom up rather than through top-down CIO mandates.
“If the value prop is that clinicians are obsessed with this tool and love it — more of like a Doximity, OpenEvidence, Heidi — that’s where we want to be. And sure, it works well and integrates, but really these doctors use this as like an Iron Man suit for how they practice. That’s where I want to be placed,” Kelly remarked.
Kelly believes healthcare AI startups should aim to sell a product that doesn’t just work well with EHRs, but also stands on its own as a high-utility assistant for clinicians and provider organizations.
To him, technology must be both functional and beloved to earn long-lasting adoption.
Photo: dan, Getty Images
