Fundamentally what it starts with is can I just understand more about these patients. So I'll give you an example heart failure, very prevalent disease sadly in this country.

Preserved ejection fraction, which is a type of heart failure, affects about 1 % of the population. But it's a very difficult disease to diagnose and often goes either undiagnosed or is diagnosed very late. And yet, if you can use both incredibly high quality data and also AI algorithms layered on top of that, whether it's through ECGs or EMR or echocardiograms, you can often find these patients far earlier in their journey.

And what's great is if you can find them earlier in their journey, yes, you can treat them, but also from a clinical trial perspective, you're finding the people you desperately want to help in terms of the phenotype. And so that's a big bottleneck is just not having that much clinical data on people and not being able to find them.

And so by using this combination that Dandelion has of this very deep data, but this AI enabled platform on top, we are able to complete that journey, find those patients, get them into trials earlier, develop the treatments faster, and hopefully cure and aid and treat many, many more people in thehave been fortunate enough to partner with the American Heart Association in the cardiovascular frame to really come together to say, you know, we believe there are AI led treatments or diagnostics out there that can really help to speed up and improve the standard of care. And so what we need to do to help that adoption is to help, you know, the patients and providers have comfort and understanding over those capabilities, specifically the AI algorithms, so that they can kind of deploy them with trust. And what we therefore do is we've created an AI assessment lab. That assessment lab assesses algorithms in terms of their performance and bias to make sure that they really do perform the way they say, make sure they don't bias against any aspects of the population. And then also has a care and economic component so that you can say to the hospital system, well, this is going to be the performance of this algorithm on your patient population. These will be, know, do you make or lose money? Do patients end up being diagnosed earlier or later? Are there any risks of full positive? But really to give an overview and a transparency to these algorithms. 

And I think with cardiovascular as the example, there's so much opportunity here to really raise the quality of care by using AI as a tool alongside physicians and ensure that they're getting into those new paths.