Latest / CareTalk: Healthcare. Unfiltered. / How AI Agents Cure Healthcare's Admin Burdens w/ Chetan Reddy, Founder & CEO, Confido Health

Transcript

How AI Agents Cure Healthcare's Admin Burdens w/ Chetan Reddy, Founder & CEO, Confido Health · CareTalk: Healthcare. Unfiltered.

0:00
David: Healthcare's phone lines are overwhelmed.
0:02
David: Patients wait on hold.
0:04
David: Staff juggle impossible.
0:05
David: Call volumes and appointments just slip through the cracks.
0:08
David: The result, frustrated patients, burned out teams and revenue left on the table.
0:14
David: But what if every call could be answered instantly?
0:17
David: Every patient need resolved without overwhelming your staff.
0:28
David: Welcome to Care Talk Executive features a series where we spotlight innovative companies and leaders working to advance the healthcare field.
0:34
David: I'm David Williams, president of Health Business Group, and my guest today is Chetan Reddy.
0:39
David: He is the CEO Confido Health, which is using AI voice agents to transform how healthcare organizations handle patient communications.
0:48
David: Chetan, welcome to Care Talk.
0:51
David: Thank you so much for having me,
0:53
Chetan: David, and thanks for the nice introduction.
0:56
David: Yeah, it's really wonderful to have you here and I wanna see if you could set the stage by, uh, letting me know, you know, what are patients and staff experiencing on the phones these days and why is access under strain at the moment?
1:11
Chetan: Access has always been limited, uh, but somehow most healthcare organizations felt like if we just staff.
1:21
Chetan: Staff, like a friend desk with people like that's gonna be taken care of.
1:26
Chetan: Uh, but over a period of time, after doing that for like almost 10 years now, most of them have realized that it is very hard to find staff who could always be available for patients and different technologies that they have tried, have not been able to solve for that.
1:43
Chetan: Uh, and suddenly when AI becomes like main topic of discussion across the world, uh, healthcare organizations have realized that maybe this is the savior now, uh, and they're just trying to go solve for it.
1:57
Chetan: But yeah, the fundamentally what's broken is not having enough manpower to take care of different patient requests that come through every day.
2:06
Chetan: Uh, and that's the core issue, uh, in focus today.
2:10
David: Got it.
2:10
David: So, you know, I've always been on the one side of the, of the counter coming in, checking in, providing my insurance information or whatever.
2:16
David: But what's it like on the other side of the desk?
2:19
David: What, what's a typical day like?
2:21
David: What, what kind of volume is someone expected to handle?
2:25
Chetan: Purely in terms of like the volume.
2:28
Chetan: I'll give you an example of one of our customers, uh, based out of Dallas, uh, their nephrology group.
2:36
Chetan: They're almost looking at.
2:38
Chetan: Anywhere from like 5,000 to 6,000 calls a day.
2:42
Chetan: Uh, and that's a lot.
2:44
Chetan: Um, and this could be from anywhere, anything from like scheduling to prescription refills to follow ups to lab requests to different types, but I would say like at least 50% of them are appointment related.
3:00
Chetan: So you're looking at at least 5,000 to 6,000 calls.
3:04
Chetan: Per day for a 30 40 location nephrology practice.
3:08
David: Got it.
3:09
David: So you know, obviously when you have that volume of of calls or frankly even something a lot lower, some are probably not gonna get through and some are gonna be unanswered.
3:20
David: And I know I've been on the other side of that as well.
3:22
David: Try to make a call to a physician's office.
3:24
David: You're on hold, you get cut off, whatever.
3:27
David: Well, what happens?
3:28
David: I mean, what's the downstream impact?
3:31
David: On patient care And what's the financial impact to a practice if a call doesn't get answered?
3:36
Chetan: Yeah, un unfortunately, uh, that's like the status quo right now of most of the specialty care practices, uh, running sizable operations today for patient access and patient, like any operations related to patient engagement.
3:54
Chetan: Um, and the biggest, I would say.
3:59
Chetan: Bottleneck for the patient.
4:00
Chetan: On the other side is the fact that there's no one picking up their phone call or they're put on hold for like more than five to 10 minutes in like many cases, depending on what's the bandwidth that's available on the other side.
4:16
Chetan: And that whole time could be really frustrating depending on why you are wanting to talk to that practice.
4:23
Chetan: Uh, and I'll give you an example of, uh, one of these patients that I was speaking to during the early days of building this company.
4:31
Chetan: Uh, patient had like a small kid, uh, and they were all sleeping.
4:36
Chetan: It's midnight, and the, the kid just started crying at like
4:42
about 1: 00 AM in the night.
4:44
about 1: Uh.
4:46
about 1: The parents were like, really, really like, uh, at that point, like really under stress.
4:52
about 1: Uh, they just didn't know what to do.
4:54
about 1: And the first thing that they would do is just call the office.
4:57
about 1: Uh, they don't know if it's an emergency.
5:00
about 1: Uh, and most people don't wanna just like get into er not knowing the emergency because it's like really expensive.
5:07
about 1: Uh, but they still wanna talk to a provider and it took them six to seven minutes before.
5:14
about 1: They got like a person to pick up that phone call and, and it's so sad that even the staff who picked up would just say that, Hey, uh, I've gotten your request.
5:25
about 1: We'll arrange a call back.
5:27
about 1: And they waited for like another 30 minutes just waiting for someone to call back.
5:31
about 1: And by then it's like 35 minutes of your child crying at like 1:00 AM in the night.
5:36
about 1: Imagine the amount of.
5:38
about 1: Uh, like stress that would, that the parents would've been on during that time.
5:43
about 1: Uh, and I feel that's the qualitative, I would say description of like, how does it really affect patients and, and financial numbers and all of that.
5:53
about 1: They're great.
5:54
about 1: It's important for practice operations, but fundamentally what bothers us is the fact that like, why do patients have to go through something like that if there's a better way to handle it?
6:04
about 1: Uh, but yeah, hope that makes sense.
6:08
David: No, it sure does.
6:09
David: And I could, I could see that it's like a literal nightmare, uh, up in the middle of the night and trying to be, uh, in touch with a healthcare provider.
6:17
David: Now, one of the things that we hear a lot about is workforce shortages Throughout healthcare, we hear about doctor shortages, patient shortages, uh.
6:25
David: Not patient shortages or nurse shortages, you know, all, all sorts of issues that are out there.
6:29
David: Where does the workforce issue manifest itself when we're talking about patient access and, and phone operations?
6:35
David: Do we just lack the people to be answering the phone?
6:38
David: Did they burn out?
6:39
David: What's, what's the issue there?
6:41
Chetan: I think the issues are threefold here.
6:45
Chetan: Uh, while there are different statistics of the workforce shortage right now on the internet, uh, but the bottom line is that.
6:54
Chetan: No practices have enough staff to handle different operations related to patients.
7:00
Chetan: Uh, but I think the reason why I call it threefold is that maybe it starts there.
7:05
Chetan: It starts from the fact that you do not have enough workforce, uh, for especially like running your call center operations.
7:14
Chetan: Uh, and.
7:16
Chetan: You may find people, but I think in many, like in many like geographies in the US uh, a person working at Starbucks like makes slightly more money than a lot of people working in healthcare and, and a lot of young generation, I think.
7:34
Chetan: They just don't want to be doing work that obviously Payless and it's such a thankless job, like taking care of patients or the day just takes so much to be doing that and like huge respect to like people who are doing that today.
7:49
Chetan: Uh, so I think that workforce shortage and the like competitive like wages versus like the other options that they have, I think that's where it starts.
7:57
Chetan: And the second issue is that, uh.
7:59
Chetan: I would, I mean, it's, it's not like a bias that I have, but I genuinely feel like the technology stack that exists today to support them is not up to the mark.
8:10
Chetan: All they have is a bunch of phone systems.
8:13
Chetan: Where there different call logics set up to queue people but not actually solve for the core issue of like helping them answer what they're looking for.
8:24
Chetan: I think that's pretty suboptimal at most of this practices and that was the best option.
8:29
Chetan: So I'm not blaming them, but that's the best option.
8:31
Chetan: So they just went ahead with it.
8:33
Chetan: Uh, and the third problem having worked with like.
8:38
Chetan: Thousands of sites right now, I can tell you that most practices do not have the right SOP or like guiding sort of materials in order to like scale that operations really well without that breaking.
8:51
Chetan: So people try different sorts of things from onshore call centers to hiring people at the site.
8:57
Chetan: To offshore call centers.
8:59
Chetan: They experiment more all the time.
9:00
Chetan: They just, none of them have like really good processes and SOPs, at least in specialty care, in lot of group practices to like absorb any sort of technology or process intervention to make that better.
9:14
Chetan: Uh, so I think these three are reasons I would say, I would point out as like core issues right now.
9:21
David: So Shayan, you've, you have characterized the problems.
9:24
David: Quite well.
9:26
David: And of course you've built up a company that's gonna address those.
9:28
David: So maybe you can talk a little bit about what Confido Health does, and in particular, you know, in the introduction I talked about handling every call, what does that mean?
9:39
David: Handling every call, uh, presumably from, you know, beginning to
9:44
Chetan: end when we say like, we handle every call, what we mean by that is.
9:51
Chetan: We are building a technology that can truly become the first point of contact for all the specialty care practices across the us.
10:00
Chetan: And when I say the first point of contact, I'm referring to a lot of inbound calls that every practice gets from all the patient panel that they have.
10:09
Chetan: It could be for like any number of.
10:12
Chetan: Reasons or use cases for which patients end up calling a practice.
10:17
Chetan: Uh, and it could also be for any sort of outbound calls for specific campaigns or specific workflows for which they wanna like reach out to patients.
10:27
Chetan: So we are building the technology and we are building a lot of processes around it to ensure every practice now has like an extended.
10:38
Chetan: Uh, agent Network that can handle both outgoing and incoming calls for them.
10:44
Chetan: Uh, however, uh, a caveat to that, the stage at which we are right now, uh, the technology cannot handle all a hundred percent of calls.
10:54
Chetan: Particularly a lot of clinical questions is something that we trash or forward to one of the staff members.
11:02
Chetan: So we still are dependent.
11:04
Chetan: On some of the staff members to be handling complex queries where there's a lot more human intervention.
11:11
Chetan: Uh, but the starting point for us has been maybe start with a lot of routine phone calls with the help of agent network.
11:17
Chetan: As we get more and more data, we build towards handling a lot more complex calls and eventually a lot of clinical calls as well, so that by then if we're the first point of contact, patients get used to this experience.
11:28
Chetan: And then now we have like all a hundred percent calls going through our agents.
11:32
Chetan: So that's the eventual vision.
11:34
Chetan: Right now we do have ability to handle 60 to 70% of calls of different use cases, from point management to refills to payments, to follow ups to certain care gaps and other use cases depending upon the specialty.
11:49
David: As you mentioned, this problem isn't a new one.
11:51
David: The phones have been overwhelmed in physician specialty offices for for a while, and there's other approaches that have been tried including technology approaches like interactive voice response IVR or this offshore call centers.
12:04
David: How different is what you do?
12:07
Chetan: The way that we look at and the way that a lot of our customers look at this technology right now is just like another.
12:18
Chetan: Onshore sort of team, but fully autonomous, does not need to go through a lot of HR policies, hiring, firing, uh, constraint, micromanagement, and all of that, just like an independent, autonomous, uh, set of agents who are onshore, who are available 24 hours seven days a week.
12:43
Chetan: Uh, because, and, and you can speak like different languages depending on what kind of patient panel that you have.
12:50
Chetan: Uh, so I think our positioning in their head is that it's cheaper than onshore contact center team and less stressful to manage because obviously, uh, it's technology and it, it's much harder to ramp up people, uh, in such sort of a setting.
13:08
Chetan: Uh, and also.
13:10
Chetan: Maybe almost equally capable of like, handle calls in a similar way, like just like an onshore contact center team.
13:18
Chetan: Uh, we don't get compared to an offshore contact center team as much because.
13:24
Chetan: Offshore contact center team typically is most like working on a lot of overflows that onshore teams are not able to, and onshore always becomes like a core integral part of your contact center operations.
13:37
Chetan: And that's what we are compared to because we are high value, a lot more scalable and very efficient like onshore teams.
13:45
Chetan: But over a period of time, we feel like all the phone systems, offshore teams.
13:49
Chetan: It would get eliminated, uh, if for the, sorry for the lack of like, good word to explain that.
13:58
Chetan: But, uh, if, if there's one vendor who could handle all the operations onshore, a lot more scalability.
14:05
Chetan: I don't see the need and customers are not seeing the need to have work with phone systems and also offshore teams anymore.
14:12
David: I'm wondering why you're focusing specifically on voice ai rather than the approach that some others are taking, like chat bots or some sort of other digital channels.
14:22
David: Why the voice?
14:25
Chetan: That's a good question.
14:26
Chetan: That reminds me of, uh, the early discovery phase of how we ended up deciding to start this company.
14:34
Chetan: Like we met almost like thousand plus.
14:38
Chetan: Providers across the US like travel to different, different parts of the country.
14:42
Chetan: Just trying to, uh, uncover the mystery behind this sort of a problem that I saw as a personal experience.
14:50
Chetan: Uh, and after speaking to so many providers, I think it was kind of obvious that, uh, most of them obviously had staffing issues.
14:59
Chetan: And in that, when we tried to map what are different pain points for which they're spending a lot of time.
15:05
Chetan: Like executing on almost four to 4.5 hours of an administrative worker typically, uh, just goes towards like phone calls and we just want to solve for something that takes almost 50% of the day-to-day time at the office, uh, so that we could show enough impact.
15:25
Chetan: Uh, for this practices, which could help them in many ways.
15:29
Chetan: Uh, hence we decided to start with like phone calls as a use case to show that impact.
15:35
Chetan: And once we become a trusted vendor, obviously expand that into a bunch of other workflows that could also be like, meaningful enough for them to continue working with us.
15:46
Chetan: Great.
15:47
David: What do you find in terms of the most common use cases?
15:50
David: I know back in the day, a lot of it had to do with, uh, refilling.
15:54
David: Prescriptions or really renewing prescriptions after the refills had had run out.
15:58
David: Someone trying to get their test results, the billing questions, scheduling, you know, what are the big use cases these days?
16:03
David: Why people are picking up the phone and trying to reach a human.
16:07
Chetan: Yeah, I think, um, at least with like many customers that we have.
16:12
Chetan: Uh, 50% of them, I would still put them into the front office bucket.
16:18
Chetan: That's largely about appointment management.
16:22
Chetan: Uh, and that's mostly true for all the specialty care practices that we work with.
16:28
Chetan: Uh, and the second largest use case.
16:31
Chetan: Uh, would be, again, in many cases about prescription refills.
16:35
Chetan: Uh, they also slightly longer calls.
16:38
Chetan: There's a lot of information gathering that happens, uh, which is kind of very routine in nature because end of the day, a provider has to sign off and providers not on the call.
16:49
Chetan: So it's mostly like the staff.
16:52
Chetan: Uh, collecting that information, uh, and then sending the request to the provider.
16:56
Chetan: Uh, so it's still like longer calls, but routine calls and mostly like information gathering.
17:02
Chetan: Uh, so we have seen like.
17:04
Chetan: A lot of calls are for refills.
17:07
Chetan: Uh, and the third most would maybe would be around like referral management.
17:12
Chetan: Uh, because all of them are specialty practices, they get a lot of refills from like other providers.
17:17
Chetan: Uh, so there's sort of coordination that's going on about that.
17:21
Chetan: Um, and a bunch of other workflows, um, also contribute.
17:25
Chetan: But I, I would say like these are like top three use cases for which like, uh, patients end up calling.
17:31
David: Well, you've done a great job of describing the problem and the solution.
17:35
David: Let's go ahead and play a sample, patient call so that listeners can hear Fido Health in action.
17:41
AI: Sarah calling from Coastal Wellness.
17:43
AI: We're reaching out to schedule your annual wellness visit.
17:46
AI: It's a really good way to stay on top of things, and we have a few openings next week.
17:51
AI: Does that sound like something you'd be interested in chatting about?
17:53
Chetan: Um, I'm sorry.
17:55
Chetan: Uh, I don't think I've ever gotten this done like.
18:00
Chetan: What, what is this for you?
18:01
Chetan: Sorry.
18:03
AI: No worries at all.
18:05
AI: That's a comprehensive checkup.
18:07
AI: Things like blood pressure, cholesterol, and a general health assessment.
18:14
AI: Basically, it's a way to proactively identify any potential issues before they become bigger problems, and a lot of people find, it gives them peace of mind.
18:23
Chetan: Oh, okay, is this important for me?
18:25
Chetan: I'm like, I'm like 30-year-old.
18:27
Chetan: Like, does it help me in any way?
18:31
AI: Definitely even in your thirties, preventative care is huge.
18:35
AI: It's a good time to establish a baseline for your health, and honestly, things change.
18:40
AI: We can check for things like early signs of heart disease or even just vitamin deficiencies that could be impacting your energy levels.
18:48
AI: It's about being proactive, not reactive.
18:52
Chetan: Okay, so that's my insurance cover for this, or am I expected to pay for this myself?
18:59
Chetan: Good question.
19:01
AI: Most insurance plans cover annual wellness visits fully, especially when they're preventative.
19:07
AI: We can quickly verify your coverage with your provider beforehand, so you won't have any surprise bills.
19:13
AI: Sure.
19:13
AI: Would you like me to do that for you?
19:14
Chetan: Uh, we can do that later, but do you have like a morning slot available, uh, next week for this?
19:21
Chetan: Uh, we can figure out the time later, but just let me know if the morning slots are available for this.
19:26
AI: Yep, we do.
19:27
AI: Um, we have a couple of openings Tuesday and Wednesday mornings.
19:31
AI: Would either of those days work for you?
19:33
AI: We're pretty flexible, so if neither of those fit, we can explore other
19:37
David: options too.
19:37
David: No, I think we can make that work.
19:39
David: Thank you so much for calling.
19:40
David: I'll reach out to you again.
19:42
David: So, Chichen, what, what did we just hear?
19:44
David: And in particular, I'm curious, how was the agent able to reach resolution without having to send that caller to voicemail or to a human?
19:52
Chetan: Oh, I, I was just waiting for you to ask this question since the beginning, but I think one of the core technology piece that we are building for is, uh, in, in typical software language, it's called plumbing.
20:07
Chetan: Uh, in healthcare, uh, I think it's lot, it's, it's a much bigger problem trying to.
20:15
Chetan: Sort of set up connections between AI and the core system of records, which are, which could be anything from your EHR to practice management, to your phone systems, to any other third party applications that power the operations today at this practices.
20:31
Chetan: Uh, I think.
20:33
Chetan: If you ask us like, where have we spent most amount of our time?
20:36
Chetan: Is it on the voice or is it on this plumbing piece?
20:40
Chetan: Uh, I would say like, uh, the latter and the main reason for that is.
20:46
Chetan: If you think of like a new staff member who is joining, like a practice, uh, staff member could only be efficient if they know how to use EHR and if they can make sense of the information that's already in the r.
21:01
Chetan: If you can't make sense of the information that you see on the EHR or any other system of record, uh, when a patient calls, you just don't know what to say, you're just gonna be like, Hey.
21:12
Chetan: Just send me the question I'll get back to, I'll give you a call back.
21:15
Chetan: Uh, which is again, not truly solving the problem.
21:18
Chetan: So we've spent a lot of time investing, building into a lot of like unique technology that helps us, uh, make those connections, get like from millions of data points, extract.
21:32
Chetan: Useful information during a call between an AI agent and a patient almost with like real time latency.
21:41
Chetan: Uh, that's what helps us run a lot of successful resolutions.
21:45
Chetan: Uh, and we, we have taken a lot of time to get that technology to that stage and we are really proud of the team that has, uh, built.
21:55
Chetan: Uh, but yeah, I think, uh, the example clearly talks about.
21:59
Chetan: How, how AI was able to know that this patient is due for a visit, made sense of the information during the call, and then was able to successfully close the loop without any human intervention.
22:11
Chetan: And there you go.
22:12
Chetan: You already saved like minutes worth conversation time for a particular staff member on that patient.
22:19
Chetan: And in this case, AI can also do that for almost 10,000 patients at once, which would've taken many hours of staff time, which they don't have to spend anymore and spend more time with patients.
22:29
David: So I'm wondering how the system knows when it needs to escalate to a human compared to just handling it autonomously.
22:35
David: Obviously we heard it handle it well there, but I, I imagine sometimes it's gonna be an issue.
22:39
David: And let me draw that out a little bit further, because one of the things that's notable when you use a.
22:45
David: Generative ai, LLM model.
22:48
David: Um, it never says, I don't really know.
22:50
David: I'll have to ask a person or look that up.
22:52
David: You know, it's just sort of like a general disclaimer at the bottom, but here it would seem to be quite important to know.
22:57
David: Yep.
22:57
David: I'm confident of that and it's gonna be correct, or I really need to escalate it.
23:02
David: How, how does the system deal with that?
23:04
Chetan: Yeah, I mean, uh, if I were to explain this to like a layman, I would just simply say that we are able to do a lot of quality check.
23:15
Chetan: While the call is happening with a lot of guardrails in place, that that is like a, like, almost like hundreds of rule sets and use cases that we have trained AI on so that, uh, it's guardrail enough to only answer questions that's confident about and then forward the calls to, uh, the relevant staff member.
23:35
Chetan: If it doesn't know the answer.
23:37
Chetan: Uh, if we were to explain this a little more technically, uh, so we also built.
23:42
Chetan: Um, we invested a lot into observability and evaluations, uh, which is basically like a way to, uh, sort of, uh, understand pre calls during calls and post calls, uh, different different sort of scenarios as to how AI should be handling some use cases and how AI ends up handling those use cases and constantly retrain the models with that kind of information so that it gets.
24:12
Chetan: Better and better over a period of time, which is where we are right now.
24:15
Chetan: So hence we have like of the best successful call ratio, uh, in this space right now.
24:22
Chetan: Uh, and that helps us like really take care of, sort of like, focus on the question and only give the information that patient is asking for.
24:31
Chetan: And like simply admit that you don't know, and then just forward the call to one of the staff members.
24:36
Chetan: And the process of like transfer is also called like a warm handoff.
24:40
Chetan: So we just simply connect the patient to the staff.
24:42
Chetan: We give staff the context, and the AI just drops off from that call, from the conference call.
24:48
Chetan: Uh, but yeah, to answer your question, a lot of, in, a lot of like.
24:52
Chetan: Our technology on observability and evals, uh, which is almost like a real time qc, that models are running on top of models that helps us do this successfully across different call types.
25:04
David: Now when the patients are calling in, I asked you a question before about, you know, what are they calling for a prescription renewal appointment?
25:10
David: And so on and so forth.
25:11
David: But there's also some characteristics of the patients that the system has to handle.
25:15
David: And I'm thinking about, you know, a practice like you're describing with, you know, getting thousands of calls.
25:20
David: Not everyone's gonna speak English, not everyone's gonna have the same accents.
25:25
David: Um, how does the system handle that?
25:27
David: Or does it?
25:29
Chetan: Yeah, I think, uh.
25:31
Chetan: Credit to, I would say a lot of foundational models, uh, to solve for this aspect of a conversation.
25:39
Chetan: Uh, if you'd asked me this question, maybe in 2023, uh, I would've given you lot more complex answer as to how this is done.
25:49
Chetan: But thanks to generative AI and great foundational models that are out there on top of it, even we are powering a lot of our conversations on by building appropriate technology on top of it.
26:00
Chetan: Uh, so I think it's almost sawed at, uh.
26:05
Chetan: Some of this model, uh, levels could be the speech to text and text to speech, uh, models, and also the inference on the inferencing side, the core LLMs.
26:16
Chetan: Uh, this is something that gets sold at that level itself.
26:21
Chetan: Uh, trying to like, make sense of the language, the accent during a call, and then sending the information back to the model.
26:27
Chetan: And then trying to determine and then accordingly give a response.
26:31
Chetan: Uh, so we've created some pipelines that enable us for the models to be behaving in that way, and then trying to have higher success ratio, uh, during those situations or circumstances.
26:43
Chetan: Uh, but yeah, I think it's, it's pretty well handled and I would think so for most companies building in this space, it's almost like hygiene at this uh, point.
26:53
Chetan: It's not like a unique feature anymore.
26:57
David: When you're installing a system like this and, and getting it up to speed, how do you measure success?
27:03
David: Are we talking about like how many calls it could handle, how much staff time it reduces, call length and abandonment?
27:11
David: What are, what are the metrics that you measure and what are really the key performance indicators?
27:18
Chetan: Yeah, thanks, thanks for asking that.
27:20
Chetan: Um, it's, as you know, it's very important.
27:23
Chetan: And probably more important in healthcare when it comes to like truly understanding.
27:29
Chetan: How some of this works in production and what it means, uh, in real life for both patients and providers.
27:35
Chetan: It's important that we have some of this quantitative methodologies also to track the success of any new technology, such sort of sensitive environments.
27:45
Chetan: Uh, so the way that we do that is we have like an internal facing metrics, which is mostly for our engineering and delivery teams to look at.
27:55
Chetan: To understand some of, uh, some of the value that technology, uh, is building based on what they're expecting.
28:02
Chetan: And we have some external, uh, facing, uh, sort of metrics that are between, uh, the agent and the customer, uh, which is in this case a practice.
28:12
Chetan: Uh, so that they also able to track the impact that the system is able to create for them.
28:17
Chetan: Uh, on the internal side, we mostly track.
28:20
Chetan: Uh, uh, we take a lot of like patient first approach.
28:24
Chetan: Uh, so we are trying to constantly track.
28:26
Chetan: How happy or satisfied where patients at the end of each and every call, we collect a lot of feedback and so away from them.
28:34
Chetan: Uh, because if the patient is happy, mostly the practice is happy.
28:39
Chetan: Uh, but if we try to build for the practice, there's very high chance patient is not happy because most practices do not have right SOPs.
28:45
Chetan: So we truly built for patient almost like a consumer company versus like a B2B company.
28:51
Chetan: So I think call satisfaction score.
28:53
Chetan: At the end of every call is like one internal metric that we obsessively track.
28:58
Chetan: Uh, and the second would be the call forwards, which is, uh, for the workflows that AI is supposed to handle, what percentage of them AI is able to close the loop on, and for what percentage of calls like it's sending back to the staff to handle The objective is to just take more of that work over a period of time.
29:16
Chetan: So we constantly track that internally, uh, and also externally.
29:21
Chetan: I think, uh, staff time saved is absolutely important so that, uh, practices can feel that impact when they know that they're like bunch of staff members and suddenly they have more time.
29:32
Chetan: They could utilize their time for like other workflows.
29:36
Chetan: Uh, so I think that's something that they deeply care about.
29:39
Chetan: Uh, and the second important external metric is that most practices have significant attrition.
29:45
Chetan: Uh, so I think they're constantly trying to figure out should we hire more people?
29:49
Chetan: Should we fill for this because.
29:51
Chetan: There are not enough people in the pipeline.
29:53
Chetan: Uh, so they're just trying to figure out, uh, I'm my, as a practice, still able to handle the same amount of phone calls or more with the existing volume, augmented with ai.
30:04
Chetan: So that's external, uh, sort of other sort of, uh, qualitative way for them to figure out, uh, the impact of the solution.
30:13
Chetan: And lastly, for, uh, some of the practices, obviously.
30:16
Chetan: Uh, we all in healthcare, we all care about impact, uh, but practices who are really struggling with the financial performance, which constraints them from, like providing better service, they end up focusing a lot on, uh, the financial impact as well, which is trying to understand.
30:32
Chetan: Uh, so what sort of, uh, how, how, how many more encounters, uh, were they able to, uh, sort of get in post AI implementation and, and how much costs were they able to save?
30:44
Chetan: Both directly.
30:45
Chetan: Indirect, which is related to staffing costs, I think is also an important metric that they constantly look for, especially when assessing the financial impact of the solution.
30:55
David: So you're not operating in a vacuum here when you come into an office or especially a. A practice that may have, uh, multiple offices, they're gonna have some systems, they're gonna have phone systems in place, they're gonna have electronic medical record in place.
31:10
David: Where does Confido sit in that stack?
31:12
David: Or how do you play in that ecosystem?
31:15
Chetan: Yeah, that's a good question too.
31:17
Chetan: Um, I mean, the way that we think about us is like just some in invisible, uh, employees at the office.
31:28
Chetan: Uh, so we ask our practices also to like.
31:32
Chetan: Look at us as partners, employees, uh, who are trying to help them finish up a bunch of work that they need to be doing on a day-to-day basis.
31:44
Chetan: Uh, and hence we look at most softwares like EHRs or phone systems and all of that.
31:50
Chetan: Uh, purely as like a way for us to sort of access information and then act upon that information, just like how an any new employee or old employee would do.
32:02
Chetan: Uh, so use the information, act on it.
32:05
Chetan: Just make sure your operations are running seamlessly.
32:07
Chetan: That's a similar approach that we have.
32:09
Chetan: Our agents get deployed.
32:11
Chetan: We teach them how to understand this information, how to make sense, sort of the information.
32:16
Chetan: We work with the staff on some change management so that as AI is starting to work on its own, what sort of other roles and responsibilities that they can collectively brainstorm upon and then trying to take charge of.
32:28
Chetan: Uh, so we do some of the change management and understanding, uh, the existing data stack for technology.
32:34
Chetan: Uh, and yeah.
32:36
Chetan: And once we are good with all of that, once we go live.
32:39
Chetan: Uh, it's mostly about, uh, just executing the job on a day-to-day basis, just like how any employee would do.
32:45
Chetan: So that's how we look at it.
32:48
David: Jayden, the last question I'm gonna ask you has to do with privacy and quality.
32:53
David: I imagine if I'm a practice manager.
32:56
David: Provider deciding whether to bring you in or not.
32:59
David: I might have some concerns about, you know, what's the AI doing?
33:03
David: It's obviously gotta listen in order to process the question and to respond.
33:08
David: And then the kind of quality, you know, it's a different sort of expertise than, or you know, quality than you're gonna have from an individual.
33:18
David: How do you give your prospective customers confidence on those topics?
33:24
Chetan: Uh, I think that's absolutely important, especially when you are building something for healthcare.
33:32
Chetan: Uh, I know I said that, uh, and I, I, I, I, I talked about observability and evaluations last time, I think.
33:40
Chetan: Uh, so I think that's like a very important piece of technology, uh, that we have uniquely to confide or health wherein we are able to give.
33:50
Chetan: Uh, more transparency towards the agent actions when agent is performing different tasks, uh, not just for our engineers to like monitor and take actions on, depending on how the agent is performing, but also external facing for practices to constantly have access to.
34:06
Chetan: Each and every call that AI is handling each and every call transcript, uh, of, from, from those conversations, each and every action item that's logged into an EHR or practice management system that they can look at.
34:20
Chetan: So there's almost like, if not for practice, like we have a team at RN who's constantly looking at each and every small action of agent, ensuring that everything is going as per uh, different privacy and security policies that we have internally and also.
34:35
Chetan: What practices have, uh, or everything from hipaa, uh, to other important aspects of, uh, sort of how AI needs to be deployed?
34:44
Chetan: I think that human supervision, uh, combined with tech sophistication on observability and evaluations, uh, I think contributes significantly to privacy, hipaa, and bunch of other important security aspects of implementing this technology and ensuring it's successful.
35:03
Chetan: Uh, something that we've been able to do.
35:05
David: Terrific.
35:07
David: Well, that's it for another episode of Care Talk Executive Features.
35:11
David: My guest today has been Chetan Reddy.
35:13
David: He is CEO of Fido Health.
35:15
David: We've been discussing how AI voice agents are solving healthcare's patient access crisis.
35:21
David: I'm David Williams, president of Health Business Group.
35:24
David: If you like what you heard, please subscribe on your favorite podcast platform and thank you Chaan.
35:31
Chetan: Thank you so much, David.