Why Healthcare AI Projects Stall Before They Help

From the episode captions, lightly cleaned. Speakers are not labeled, and the text may contain small errors. Time links play the episode from that moment.

0:00 In medtech, the difference between a good quarter and a great one isn't always the product. It's the friction standing in the way. Broken sales cycles, misaligned teams, deals that stall at the finish line. Every leader feels it, and the best ones know how to remove it. Welcome to Zero Friction, the podcast where we go inside the minds of MedTech's top commercial leaders to uncover the real barriers they face, the decisions they make, and the systems they build to keep the revenue moving. I'm your host Terry Coutsolioutsos. Let's remove some friction. Hello Zero Friction podcast. This is Terry Cuchaluchas. I have a great guest today that's going to bring us some perspective from the platform and the AI space. I have Deepak Borole. He is the director of operations for the health care vertical at Chetu. And basically what they do in the marketplace is they are a technology developer and his group is focused on how to bring new technology, new efficiencies and now AI

To the healthcare vertical. So Deepak welcome to be here today. Can't wait to learn more about you and your your history in healthcare and what you're doing and then let's talk about AI and healthcare and work what Chetu is doing to help shape the future of that. Thank you. Glad to be here. Well, good. So, why don't you tell us a little bit about your background? How did you get started in technology? How long have you been with so about me? I am Deepak Borole. I'm the director of operations here at Chetu managing the healthcare vertical.

Something about more about me that I did my education back in India did my engineering and that's how I got in technology. I started as a developer and then went across becoming and performing other roles like team lead, project managers and then the director. Chetu, I've been working for more than 12 years now and purely focusing on healthcare. So we have been doing a lot of uh interesting and different kind of projects here in Chetu when it comes to healthcare whether it is for general healthcare, charm, remote healthcare, chel healthare and even the pharmacy and life sciences.

So you've seen a lot of changes with you know the emergence of EMR automation and now AI is on the forefront of things. So tell us about how Chetu engages health care providers today and and what are your biggest most promising offerings to healthcare in today's marketplace? Sure. So Chetu is more of like AI energy transformation company. So we help customers to build solutions and products. So we have engineers who work with customers to build their vision, build their workflows, build their solutions which they own it. So all the intellectual property is owned by the clients. So that's our model. And again um apart from that the main differentiators what uh we have is that we as a company when we work with a client we provide a team who have extensive experience in the particular domain area and we understand the standards protocols in healthcare like HIPPA PHI any other guidelines and all which we use and provide feedback and consultation to the client while working with them. I'm guessing the customer is

3:28 Having experiences with some sort of friction, right? What are the biggest friction points that you and your team are dealing with healthcare and what problems are you helping them solve today? So, uh the top most um challenges which the industry in healthcare they facing is basically the burnout problem, right? the providers, nurses and other staff members they they have so much of work task and all they don't get time to do everything right so that's where uh these systems transformation and AI is helping so where we can automate their task uh some of the routine tasks which they do on repeated basis can be automated pretty easily through AI that way some of the load can be taken off them example like ambient listening a successful thing which has been uh in the market for some time now. But previously providers or the nurses used to uh take notes and then enter them in the system. Even they used to call something called as pajama time where they used to go home and do that

Activity. So now with the help of ambient listening this can be easily recorded, translated into a proper format and then it can be integrated in the EHR. So they just need to review it once and approve it. That's it. Excellent. So you take the requirements from the customer, right? Like you figure out what the problems are. You then take an inventory of what they need in order to solve that problem. Where where are the most problematic areas of the project that you know keeps them from moving forward.

Sure. And you rightly said that we work with them end to end. We work with them to gather the requirements then plan that out into project plan and there or even the task and all execute them, deliver them and even provide support and maintenance and again obviously there there are some challenges which come across while working with uh in this particular format right sometimes the the client is not able to finalize the requirements or they keep on changing it due to which the time and the cost increases That's one of the thing sometimes they themselves are not sure about what they need and they don't want to spend time on the user acceptment testing which is very key component of any SDLC life cycle. So some of these challenge which we see while working with the customers but over the period of time we have seen lot of clients we have worked in different industries including healthcare. So we understand what questions to be asked to

6:06 The client from the get- go to capture all that. That way we can reduce the friction as much as we can. Understood. So one thing that you mentioned earlier in the conversation after the problem has been solved, you said the client owns the IP and they can white label. What does that unlock for the provider itself? Right? So basically imagine that you basically have an idea, right? And you want to build something. You're starting a business. You have an idea. Uh you want to build a product, right? So we as a company we help that particular client to build that product the way they want it. So we kind of take their vision build that product for them by closely working with them and they own the intellectual property of it. So they can uh use easily widely application to get more users subscribe and all those things can be done. Even they want to sell this product tomorrow they can do that right. So it's like offtheshelf

Software versus a custom solution which they own. Right. So tell me about the repetitive work areas in the hospital. Where are the big areas where that's found the most or the healthcare provider? Right. Some of the common use cases which you can see is that um the front staff where they get a lot of calls and again not only calls but they lot of get lot of faxes, mails, notifications right from the patients or the users right that's one of the repetitive task which that particular group has to work every day. So if if we can automate some part of the work for them they can use that time to focus on something else right at the same time what we do is we build human human in loop component which can be used for some critical scenarios or workflows. So that's one area. Other areas is documentation where uh different uh members have to document the work that they do whether it is providers, nurses

Or even the staff right that is where AI can be used to automate that. Apart from that there are also areas where they wanted to generate some reports. Sometimes they might be doing it manually. Right? They kind of I want to get some data on the fly. All right. So previously they were to create a request somebody used to work upon it and then give the feedback. So all these things can be automated AI can be leveraged to give them easy access to that data. So they can easily query response and use that data for their ongoing task. Got it. So one thing you mentioned to me before in our earlier conversation really was revolving around the medical practice itself. And so let's talk about the front desk and what you do for for practices, right? I think that's a a big part of a marketplace that has strong demand for these types of products.

9:08 Correct. Mhm. That's correct. So basically the burnout problem everybody's facing in the industry, the shortage on resources also, right? They may not get the front staff, right? Or even the sometime they need back office staff also. So back office is nothing but who does some task ahead of the schedules. Example they wanted to a generate some data and have it ready for the appointment. Vice versa they also work on billing and coding. So all that things basically are back office and the uh front staff is basically who is working with the patients. So so the idea is that is nothing but a workflow right. They do it for every patients. There are kind of scenarios which need to be covered. What we can do is we as a company can help them to understand those workflows do the automation for them by going from step one to step four and so on so forth. But the idea is that we also integrate that data with the EHR they're

Using. So what we can do is we can not only automate the task but also integrate that data with the EHR that is this end to end automation is achieved and as I said earlier that uh for some of the task we will also give them options to do human in loop where we can provide them dashboards or screens where they can review that data content and then approve it. Very good. How long does it take to you know build a platform for the front desk typically?

So I would say uh it depends on what the features they're looking for but I would say in it in the range of like months it's average I have seen uh by working with different kind of clients. Yeah. Right. So if I were to come in and say I need a front office or front desk solution. What does that process look like? Tell me a little bit about what how does it start? what happens in the middle and then what happens at the end. Sure. So what we do is we first understand what practice that is right what are their patients look like what is their workflow look like what is the ongoing things and what is their pain points right so we capture all those details right and what they're looking for they're looking for some voice integration right as I said earlier automation of uh email facts or even text messages so we capture all these details we create um design documents where we detail down everything which is required as part of this process. Then

11:41 We create a project plan which is nothing but breakdown of that scope of work which will give a clear visibility to the client. Okay, this is what that workflow or the task would be and how much time it would be and then we work upon those tasks through using that project plan and show the progress every week to the client. That way they know what's uh what's happening any questions those can be answered and we also provide them intermediate releases that way they see what's going on and don't they don't have to wait till everything is completed and we have a clear plan of testing then going live the production environment and then providing the support after that because it's a software right you will need some support on ongoing basis so we do have different offerings around that if a clients want to have continue with a smaller team we can do that but otherwise We also offer a block of uh our services or packages to support them

As part of support and maintenance. Got it. So when it's all said and done, h how does the practice trust AI today? Like you come in, you're trying to develop a new solution to create efficiency to make life better. How does that all go? And like what are the pluses and minuses when it comes to you know working with the practice? Correct. As I said earlier that they may not know exactly what they want right they just say come here okay I we want AI because they have heard about it right so we narrow it down exactly what you need like what is your pain points and then according to that we we understand their business services they provide and come up with the high value use cases which apply to them but um um in some cases we also do proof of concept right first to show them okay this this is your pain point, this is your high value use case, this is what

We will do. And once they are like on boarded with that, then you kind of expand that to scale it up and uh implement that for the full practice or even multipractice scenario. It depends on practice to practice, but it's not like it's just a product we give them. Okay, this is your product. You take it and just plug it and start using it. Got it. tell me about like billing and filing claims and like what's going on with that in the world of AI today.

Sure. So yes, there have been uh a lot of usage of AI in this particular space because um a lot of this activities are repetitive, right? The building and the coding activities are repetitive. At the same time, there's a whole process which goes uh in the background for claims, right? So basically the providers have to file claims which goes to the payers of sharing house uh which takes time and there's lot of complex data involved over there. There's a lot of denials there's back and forth even you must have heard about priorization right which takes a lot of time to get approved and so on. So all that thing can be now automated by health of AI. So what happens is that with these denials and errors can be reduced using AI to achieve the accuracy. That way the claims can be approved quickly.

14:52 Reimbursement can be done sooner. And with this especially this helps in lot of hospitals especially rural ones who already have uh challenges with the funding and expenses. So it helps them a lot. uh apart from that an inteative work which I said uh now can be done through AI and uh again it's kind of around the clock you know I heard that insurance companies have been using AI for file for claims denials as well so now we have doctor's offices fighting with the insurance claim AI so it's AI versus AI how does that how does that work I would say that overall the accuracy is improved right with the time and all would be reduced because imagine that somebody is doing it for days and then submitting it then they used to process it. So overall we are using on the provider side on the payer side also AI is using so obviously the process has become faster and the responses also so we are able to catch those issues and

Errors pretty quickly that's like how AI versus AI is working now to move faster and even the same thing like the priorization case that that has been a a huge pain point in the industry so now instead of taking a lot of days to approve that and then keep the patient waiting or even when admitted in the hospital which increased the cost right so now that is reduced a lot so it it process through AI and also the the on the other side there's an AI which who checks a lot of things and then approves it understood and then tell me about your integrated decision support and how does that work in the EHR the EMR these days from a workflow perspective tell me what's important there.

So yes in the EHR world right every hospital every clinics have been using EHRs. So EHR themselves are having some AI features in the platform agents which are commonly used across uh different providers and um practices and all but there are still areas where uh AI is not there or there is some custom thing which a practice need that's where we come into play. We help them to build that AI and also integrate that data which is required for the AI into the EHR through standard like HL7, FHI and so on. So we use that extensively to do the end to end integration and build that AI component for our clients based on what they need and they can then use it or even enhance it as they need. I come from an imaging background space and I know there's a lot of AI used in imaging. So tell me how AI is helping to automate the clinical decision making clinical workflows in the imaging space because I think that's an area that has

17:47 A tremendous opportunity for efficiency as probably as well as quality. Correct. Yeah. Again imaging has been used a lot obviously everywhere in healthcare. So a lot of scans happens uh every day right a lot of soologist especially has to go through a lot of scans they have to work on their side and give the results back right so all that thing now can be automated through AI so example we're not saying like replacing the radologies but the common things which they do example if they are um going through scans and uh categorizing the different areas right so what we can do is through AI Yeah, we can uh do that and categorize those scans into different buckets or different problems like example this is for pneumonia this is for this area and all that thing that way uh to 80% of work of just can be automated and they can then they take a look at the end and then work on other areas. So in short

They can process more scans in a day which they used to do earlier. At the same time if a human is involved error can be there right uh there can be negligence and so on with AI it's more of like accuracy so the more better data which we have to train the models and all the better is the accuracy so obviously data plays a very important role uh and we have been working on a lot of such projects where we have taken lot of such images or scans to train the models and get the accuracy or even fine-tune the models to uh achieve it At the end of the day result is that uh it improves the accuracy and also the time for the providers or urologist.

Got it. So if you're a healthcare leader where do you start? How do you even initiate an AI engagement process right? So yeah so if you are a provider right so I I would say that in some cases just know about AI like heard about AI they don't know where to start. In some cases they know what exactly they want. Right. But the the idea is that most of the companies are what I've seen they have been focusing upon working with vendor like us. Right. or in some case they also have one or two resources in house but it's a combination or hybrid approach where they use a vendor like Chetu plus uh internal team to work on their um high value use cases uh maybe do a proof of concept right and then uh expand that and scale it up and then move on to the other use cases that way they start getting the outcome and then they see other areas where AI can be used to automate things.

20:34 Got it. So, of everything that Chetu builds, where do you think in your world in the health care world where's the fastest payoff when they invest in automation? So uh I would say it is the time and money right for the healthcare providers if you imagine that if if they had some five people working on a task on a repetitive basis now that can be reduced to one one person job who is kind of more of like doing human loop right so that's where the time is reduced and also at the same time they are paying those resources the money is also saved in the way but the the good part is that those so same resources can be doing some other work.

So that's what value I've seen on all across different projects or different clients. Got it. And what's the first friction point you think they should tackle like the most repetitive task that they have? I mean we're if you put like your problems in a box, right? Like in four boxes. So you have high impact repetitive things, you have low impact repetitive things, you have high impact not repetitive and low impact. So focus on the high impact or focus on the high activity, high residual activities. Is that where they you you tend to have your customers start?

Correct. Correct. So I would say it's a combination depending on which area it is that it is clinical versus administrative. Right. If it's like administrative, if it's high impact and operative, obviously that should be the way to go. Sometimes clinical need to be taken carefully because it deals with patients and also maybe that can be dealt with after we do the administrative task because now we know this is automated. So now there's focus upon the other clinical use cases where we can do automation but may not be 100%. Maybe 60%. That's totally fine right because we have to be careful at the same time we have to follow the compliance and the governance we have to put some guard rails that way uh we get the required output at the same time health uh is maintained of the patients uh the providers are dealing with especially in healthcare with administrative burdens I just think that there's so much opportunity in the physician office space how much work do

You do for physician offices today a lot I would say a lot a lot of our customers are physician offices okay providers um um who who can be um single practice or multipractice facility but yes they do a lot of have lot of needs around um modernization or transformation where they're doing a lot of manual tasks where we help them to transform it or even automation integrations with the different platforms not only EHR but they might have other systems also which are working in silos right so they need help to talk to each other and that also help them solve a lot of their problems so tell me are there any other points you want to make to the audience before we kind of start closing things up yes as I said that we have been working in this space for last 26 years so we have extensive experience in healthcare and um there's a lot happening uh with AI and healthcare lot of uh scope to automate uh and uh uh modernize things

23:53 Uh try to uh work on those areas and work with company like us who can help you to achieve that. Good. So one last question for you know Deepak if you could go back to your younger self coming out of uni or university right what would be the one piece of advice that you would give yourself? I would say um for for people who are passing out right my message would be be curious right be hungry and have adaptability because uh things have changed so fast and keep on learning right because that's the way to go otherwise you will easily fall behind well Deepak I appreciate you being here today if anybody in our audience Chetu is a bespoke developer for health care practices they have a big portfol portfolio, but Deepak runs the healthcare vertical. And if you have a doctor's office or a practice that is looking for front desk support, administrative support, EHR handling, claim filing, all those things kind of fall into the

Umbrella where AI can really create some value for our practice. Correct. That's correct. All right. Excellent. Thank you, Deepak. Great to see you today. I appreciate it. Thank you. That's a wrap on Zero Friction, the show built around one simple idea. Remove the barriers and the revenue will follow. If today's conversation was valuable, please subscribe and share it with somebody who needs to hear it. That's how we will grow this community. I'm Terry Coutsolioutsos, revenue architect powered by AI. Come find me at my website, myrevrx.com, and let's talk about what friction is costing you and your organization. Until next time, keep removing the barriers and keep the revenue flowing.