So can you unpack that foundation for us a little bit?

Yeah, and by the way, you nailed it there, right? If you try and bring AI and automation to a broken system, it's going to exacerbate the broken system.

So I hope all the practitioners here understand that.

And it's like really taking an honest look at what you have and what kind of systems you have.

So one of the things that we are seeing in the healthcare system is, especially with the larger enterprises, is one is they've got a whole disparate number of systems in place right because what you've done is over the last 10 years 20 years you you've acquired a lot of assets and as you acquire assets they might be on different systems of records so when we go in and look and do the first initial audit we are first just trying to figure out like okay let's let's take a look at how many candidates did you get in the last 18 to 24 months Let's laser focus on that cohort versus trying to boil the ocean and going back five years, six years, 10 years where the data is really noisy and bad.

Because if you try and layer automation on the bad data, you are going to get lots of spam going out with very low quality signal going out.

So we really tend to focus on the 18 to 24 months.

When you all look at designing, and you'll hear this term thrown around a lot, but this omni-channel, so multiple channels for clinician, can you talk to us a little bit about the differences in engagement there and what determines what next touches should be, whether it's text or voice or chatbot or human recruiter? Is that all engineered in your system or is it unique against those bifurcated roles that we had talked about earlier?