Chetu blends human ingenuity with cutting-edge AI to build custom software that transforms productivity, accelerates growth, and unlocks measurable business value across 40+ industries. With 25+ years of innovation, global delivery teams, and enterprise-grade security, Chetu delivers intelligent, tailor-made solutions that help organizations modernize with speed, precision, and confidence.
In this episode, we speak with Deepak Borole, Director of Operations at Chetu, about the integration of AI and automation in healthcare software development. We discuss how AI is being woven into the development process to enhance efficiency and compliance, particularly through Chetu’s “Attract to AI” approach. This method aims to streamline coding, improve quality assurance, and speed up delivery while maintaining strict privacy standards. The conversation also explores the future of agentic AI in healthcare, emphasizing its potential to support clinicians, improve patient outcomes, and optimize operational processes like billing and scheduling. Deepak highlights the importance of custom solutions over off-the-shelf products for specific healthcare needs, ensuring compliance and maximizing ROI.
Watch the podcast here:
Listen to audio only here:
Learn more about Chetu.
Interested in appearing on the SourceForge Podcast? Contact us here.
Show Notes
Takeaways
- AI is revolutionizing healthcare software development.
- Agentic AI can make autonomous decisions in healthcare.
- Custom solutions provide flexibility and ownership to clients.
- AI can significantly reduce billing denials.
- Compliance with HIPAA is essential in AI implementations.
- Remote patient monitoring enhances patient care.
- AI improves operational efficiency in healthcare settings.
- The future of AI includes microagents for specific tasks.
- AI supports healthcare providers, reducing burnout.
- Measuring ROI is crucial for successful AI integration.
Chapters
00:00 – Introduction to AI in Healthcare Software Development
02:50 – Understanding Chetu and Agentic AI
05:48 – The Evolution of AI in Healthcare
09:04 – The Role of Agentic AI in Healthcare
11:44 – The Future of AI in Healthcare
14:47 – AI’s Impact on Healthcare Staffing and Provider Support
17:55 – Ensuring Compliance and Governance in AI
20:45 – Remote Patient Monitoring and AI Integration
24:06 – Enhancing Patient Experience with AI
26:58 – AI in Operational Efficiency and Billing
30:02 – Integrating AI into Existing Healthcare Systems
32:40 – Custom Solutions vs. Off-the-Shelf Software
35:47 – Measuring ROI in AI Implementations
38:42 – Case Studies of Successful AI Implementations
41:49 – Conclusion and Future Engagement
Transcript
Beau Hamilton (00:00.654)
Hello everyone and welcome to the SourceForge Podcast. I’m your host, Beau Hamilton. Today we’re taking a closer look at how AI and automation are changing the way healthcare software gets built. I’m sitting down with Deepak Borole, Director of Operations at Chetu, a company that’s been building custom healthcare solutions for two and a half decades now. Everything from electronic health records and telemedicine platforms to billing, imaging, and remote patient monitoring.
I think what’s especially interesting is how they’ve actually started weaving AI directly into the development process itself. It’s 2025, it’s almost 2026. And so if you’re not implementing this technology into your platform, you’re putting yourself, I think, at a really serious disadvantage and leaving a lot on the table. And I think the key is really implementing this technology in a thoughtful way that doesn’t add more confusion and provides real value while also, of course, following the strict privacy and compliance regulations that are in this industry.
So the approach Chetu has come up with is actually called “Attract to AI”, which gives teams a structured way to bring in AI into their healthcare products without all the usual friction or long learning curves associated with that endeavor. So in this conversation, we’re gonna unpack some more and hear about how their AI tools help streamline coding, do all sorts of things like improve QA, speed up delivery, and just make complex healthcare builds a lot more manageable.
And of course, we’re going to explore what this means for providers, for patients, and anyone trying to innovate in healthcare right now. So without further ado, let me get right into it. Deepak, welcome to the podcast. Glad you could join us.
Deepak Borole (01:38.686)
Thank you.
Beau Hamilton (01:40.31)
Now, let’s start with the basics. Let’s start at the beginning. Who is Chetu and why are they at the forefront of this agentic AI movement we’re seeing, especially here in the healthcare industry?
Deepak Borole (01:53.237)
Sure. Thank you. Thank you for having me first of all. So basically, Chetu basically is an AI and transformation company. One of the vertical that you focus upon is healthcare and life sciences. And as you rightly said that we have around 25 years of experience building different healthcare solutions, whether it is EHR, EMR, motorcycle management, or even interoperability, pharmacy solutions or behavioral health and so on. Now, along with that, we also have been spending a lot of time and effort on building AI solutions. And AI, we build not only for our customers, but we also adopting that internally into our fabric. So whether it is HR department, accounting, or whether it is using for development purpose or quality assurance, right?
Now coming to your question about agentic AI. So yes, definitely agentic AI is what basically it understands the context, right? Make decisions and take actions on the clinical and operational workflows of healthcare. So not only just the custom AI like example, just asking questions and getting answers. So this is one step ahead. So agents basically can work on its own and make decisions.
So that’s what we are focusing upon. And again, we are building different agents for different use cases. And that makes our company an implementation and innovation partner and not just a software vendor. So we are working with different providers, players, health care companies, and so on. And that’s how we are helping them to integrate and implement AI into their environment, their fabric.
Beau Hamilton (03:56.206)
Gotcha. All right, that’s a great explanation. And it’s really exciting to see this transition into this next technological area. I mean, you guys have been around for almost three decades, two and a half decades and counting. And so you’ve been through a lot of different changes. Like the internet has changed a lot in that time frame, moving to the cloud and then dealing with a lot of the kind of regulations and compliance stuff, among other things that have happened in the last two and a half decades, been a whole podcast talking about the different changes.
There’s a lot of, I think in regards to AI, I think it’s very exciting. I know there’s a lot of really smart people trying to figure out how to best utilize AI in the healthcare space. I mean, some are, I’ve seen business leaders and programmers looking to kind of capitalize on this technology. Others are the patients and the providers actually who care less about the money and just want to see firsthand areas that can be significantly improved upon it. And a lot of those pain points firsthand that can be fixed.
So my question for you is with kind of all this sort of competition and, you know, a lot of people trying to tackle the same problems. What would you say separates Chetu from other custom software and AI solution providers out there? What, sort of advantages do you have that others maybe don’t?
Deepak Borole (05:20.916)
So as I mentioned earlier, we do custom solutions. So whatever we build for our customers, we build based on the requirements, their use cases, their needs, their pain points. And one of the differentiator is all the intellectual property is owned by the client. So tomorrow they want to make any changes to the workflows, make any additions, make any enhancement. They can easily manage that, which they may not get to all the other vendors like off the shelf software because they are like tied to what features they have. So in our case, it’s custom, not cookie cutter.
At the same time, when it comes to healthcare, we focus upon compliance, right? So whatever platform we build, we not only guide the client to how to build compliance software, but work with them from start till the end and also give them the required support and at point of time if they want to enhance anything we can do that for them.
Beau Hamilton (06:28.398)
OK. Yeah, that’s very, very exciting to hear. I think that’s going to be very appealing. It’s just that custom solution set as opposed to the, like you said, the cookie cutter approach, the off the shelf sort of approach that you see as other options out there, I suppose.
One thing I’ve heard you obviously mentioned a lot is this agentic AI term that just appears more more often. I think it’s definitely the next kind of evolution of this AI revolution we’re in where you know, we typically have the the chatbots the LLMs where we interact with and ask questions and we get information But the agentic part of things is interesting because these this kind of tool and feature allows AI to go out and actually accomplish tasks for you, once it has a good understanding of what it is that it needs to set out and do and accomplish.
From your vantage point, well, I guess a lot of organizations I think, this in general is still pretty new territory for a lot of organizations. So I’m curious, like from your point of view, what does the landscape look like right now as opposed to like where it’s headed in the future? Like what kind of AI tools and services are being utilized right now as we speak?
Deepak Borole (07:43.253)
So let me kind of elaborate that, especially when it comes to healthcare, right? People are careful. Like an example, they deal with patients, their health and all, right? They don’t want to implement latest technology as it is. They want to first see how it performs and then use it, right? So they are cautious about it.
So what has been happening in healthcare is basically, so far the healthcare organizations are interested in seeing what are the kind of AI use cases they can use, what are the ROI’s and all, what are the pilot programs they can run. But this has gone throughout this year, but I can see the transition now. They are looking for more outcome focused implementation. Now they move from pilot to implementation, right? So that’s where agents come into place or what you call agentic AI. So basically, agent is nothing but an autonomous task which has clear goals, clear access to tools and data. And it can work on its own and can support the humans, support the staff in the organization, support the nurses.
So it basically works in conjunction with the clinicians, give them required data, right? Because end of day human in loop is very, important, right? Especially in healthcare. We cannot give everything in hands of agentic AI and ask them to work with the patients, right? So, but definitely with agentic AI, we are seeing different kinds of outcomes where it is a denial reduction when it comes to billing, reducing no shows. And then also improve in compliances, majors and all. So definitely there are different areas where we can see the outcome of it.
Beau Hamilton (09:46.124)
Okay, so yeah, it’s moving to this outcome based approach or phase, because it does feel like we’re in that phase where the foundation is still being built. at the same time, we’re in that transitionary period, because I know that agentic AI is a term that I’ve really started to see it kind of being floated this last year and a half, almost two years or so. And so that makes sense, just given the natural progression of technology, that we are moving from this pilot phase into kind of a more implementational phase, where it’s like, all right, we’ve been through the testing. Let’s actually release these features and have them actually get started and work in.
So I like that timeline. I’m just curious. Maybe I’m getting ahead of myself here with this next question. But where do you see all this moving towards? Like, I don’t know if it’s, how do you predict like the next three to five years or so with this technology?
Deepak Borole (10:49.172)
So I think in the next year or so, I think we would see a lot of smaller agents, what you can call as microagents, which will focus upon areas like, for prioritization in health care or coding and documentation solutions, or even patient engagement, patient discharge and all. So that’s what you will see the next year.
In the next three years, as it will evolve, right? We can see multiple agents work in conjunction. So imagine that different agents are focusing upon what they are supposed to do, like, example, optimizing any staffing requirement or managing any throughput or giving some data to the providers which they need at the point of care. So they’ll work in conjunction. And there will be an orchestration process to manage everything around it.
At the same time, there will be guardrails, which is called as governance, right? Because that is very, very important. Without that, we can get output which will deviate from the accuracy, right? So that’s what we will see in the next three years. And beyond that, I would say that this agents would go deep into different organizations’ platforms, whether it is EHRs, or into the organizations like hospitals and all where it is continuously used for fine-tuning different departments or identifying the populations who are at risk or having any chronic care problems. So they will go deep and not just at a high level like what is happening now. So that’s what will happen in the next, I would say, three to five years.
Beau Hamilton (12:41.55)
Right. Well, it’s exciting to see how this all evolves. And I think we’re going to we’ll come back to kind of the future and what this looks like later on in this conversation. But one one thing that I want to kind of explore further is you mentioned obviously the mention of guardrails, making sure there’s there’s proper governance and then keeping a human in the loop. That’s a big that’s factor for a lot of these a lot of this technology across different industries.
A lot of people I’ve talked to, they’re emphasizing the importance of keeping a human in the loop. I think that kind of goes along with one of the fears that I hear mentioned a lot is how AI will replace human workers. And specifically, in this industry, maybe one of the concerns is these automations will inevitably lead to replacing clinicians, maybe even your doctor or your physician. Is this a valid concern of yours? if not, how do you think agentic AI will help actually maybe just support doctors and enhance their ability to prevent and treat various issues?
Deepak Borole (13:52.277)
So basically, if you see, right? There are different problems which the healthcare industry is going through. One is staffing shortages there, right? There are not many providers or nurses who are required, right? In the country. At the same time, they have a lot of work. There is a burnout because they have to do a lot of documentation. They have to spend a lot of their time in the software compared to working with the patients.
So, AI will support them. AI will help them to reduce the time on the software. For example, reduce the time on the documentation. For example, they call it pajama time, where providers go back home and they use that time to do documentation. So, that can drastically reduce with the help of AI.
Similarly, other staff like nurses and all who work on different documentation, there are being different use cases identified where AI can help them to automate it or at least help them out which they can then they review and then submit it. So all in all it will not replace the providers or the healthcare staff but will help them to move faster and spend time with the patients. So but I would say that AI will not replace all the providers but will replace providers who are not using AI. That’s what because everybody has to become smart and move faster.
Beau Hamilton (15:24.972)
Yeah. And you have these new tools, these new technologies, you have to adopt it. Otherwise you’re just, yeah, you’re just going to get left behind and you’re just going to make yourself have to just work extra hard because these tools really do help free up a lot of resources. I mean, one thing, I think that’s a good way to frame it though, because I mean, if you look, anyone has gone to the doctor’s office, at least here in the States, it’s like, never see nurses and staff just twiddling their thumbs waiting for something to do. There’s always something for them to do. There’s always more areas where things can get more efficient. And I think it goes hand in hand with the shortages. I mean, there are a lot of, you know, unfortunately, nurse shortages, doctor shortages. So if this technology can help improve that and alleviate some of that, I think that’s absolutely a win-win for everybody. And if it translates to having more time with your doctor to actually talk about your issues, that’s what it’s all about.
Now, whether a doctor is replaced by AI or not, or maybe they’re just using it to take notes or help make decisions. Whatever the use case might be, this does raise some privacy concerns, of course. I think the most potential, almost kind of obvious violation would be with HIPAA laws. How do you ensure these systems meet HIPAA requirements, but also stay within governance and don’t introduce various biases or operational decisions? How do you make sure they have followed the various regulations out there?
Deepak Borole (16:58.548)
So let me just break this question because a very, important question when it comes to healthcare. Basically, it is framed around compliance, bias, and governance. So compliance is what you said, the HIPAA compliance. So what we do is on all the projects, healthcare projects that we work upon, we make sure that the HIPAA compliance is followed and implemented. The team who are working in our company are all hyper trained who work to focus on healthcare, right? So we implement all the compliances, the logs audits that way any task which AI will perform that will be tracked, right? So that will be tracked as part of impact compliance.
Apart from that, we also have bias, right? So bias when it comes to it can be broken into like three parts. Example, what data goes in then there is a model evaluation and what are the policy constraints as in that what it is allowed to do. So if we define that very properly, then the bias can reduce.
And governance is nothing but what you allow that particular task, AI agent, to perform. If you have some guardrails put around it, it has to perform so-and-so task and not go out of it, then obviously we’ll focus on that.
So what we do is we help our customers to define that, to make them aware that this is what it will take to implement the HIPAA compliance and how to maintain the bias and how the governance guardrails need to be implemented. So we do consulting, we do have framework around that, and we help them to design and implement it and then show them the output and continuously monitor them how the output is being evolved and then how the accuracy is being met and along with those compliance are followed.
Beau Hamilton (19:04.974)
Gotcha. OK, that’s great that you’re focusing on the importance of this area and defining the variables, knowing what data is actually put into these models, and just keeping tabs on everything. It almost seems like the bias monitoring and audit trails are becoming as important as the models themselves in a lot of ways, especially in regards to, you know, when you’re dealing with patients information and healthcare and all the sensitive matters, you really got to prioritize that above everything else it seems like. So okay, that’s a great kind of explainer and it’s very reassuring, I would say.
Now, one of the many methods that continues to grow in prevalence is remote patient monitoring, which for listeners who might not be unfamiliar, it’s the use of digital devices to collect and transmit patient health data to healthcare providers for review outside of traditional in-person visits. So these are things like Apple Watches, maybe Smart Scales, blood glucose monitors. There’s a lot of different examples. But maybe you can talk about the role AI will play in this area. What will its impact be for both providers and patients and the care they receive?
Deepak Borole (20:17.716)
And as you rightly said that in remote patient monitoring, there lot of parameters which are collected from the patients, whether it is using devices or in some cases, there is software as a device also used to variantize input by the patients into software. That’s also one part of remote patient monitoring.
So how AI will help over here is to gather data from all these different sources, which will take time for the users to review and monitor. So AI can aggregate all that data, detect patterns in that and distinguish it from like example, what is the noise and what is the quality data out of it, identify trends. And then based on that, it can triage the patient’s parameters or health, identify any low risk interventions. At the same time, it can educate the patients based on that parameters or send them reminders, send them questionnaires. And if there’s a high risk, it can escalate that to the required providers. And to the right care team can be notified and they can connect with the patients.
So that’s where that it doesn’t need to be done manually. So all the AI agent can take care of all that. The same time it notifies the required providers or care team about the patient’s health.
Beau Hamilton (21:46.286)
Okay. So it’s, kind of connecting the dots, extrapolating, results, taking all that data and actually making something out of it and connecting all the different, you know, the hardware with the software. I think that’s just such a fascinating, it’s such an interesting time. I think we’re going to see some serious innovations here in the next, I mean, we already are, but the next, you know, few months and a few years, like I just found out it was kind of random, but I found out that, atrial fibrillation, common, you know, heart arrhythmia issue, it’s being increasingly diagnosed. There’s been a big spike these last couple of years, but it’s not necessarily because of, you know, healthy, like outside environmental factors or anything. It’s because people are wearing Apple watches and Apple watches are now able to detect for this.
So that’s just like one kind of example of how hardware technology is kind of shining a light and helping improve people’s health. But hearing about these AI software implementations, I think is really interesting and maybe be able to take some of those that that data from the hardware of the Apple watch and turning it into something useful I think would be really valuable. But zooming out to that patient experience. I think this next question is pretty self-explanatory but I want to still ask it to you anyway. Where do you see AI making the biggest difference for improving the patient experience and the overall health outcomes? Where do you see the biggest kind of value add?
Deepak Borole (23:14.004)
The value lies on the both the sides. As the providers and the patients side also. On the patient side, obviously the AI can track each patient separately, right? It’s not like we can have a personalized approach for each of their patients. They’re based on their history, right? Their medication. And then based on that, the education, the reminders, the referrals, telehealth, etc., can be personalized or tailored for each individual patient. So you might have heard about assistant, right? AI assistant.
So nowadays, that has been picking up. each of like there are assistant being created, which will know, okay, this is the patient, and it will know all his history. So even if patient has any questions, they can easily just go there, ask any question and they can get a response. At the same time, the AI agent will support that. So it can do some tasks in the background apart from that assistant and feed that data as part of the outcome which patient would need and have a better experience in the sense that they will have a 24/7 responsive care, right? Not limited to office hours, any follow ups, medication reminders, any care navigations, I mentioned personalized education they can have through this. At the same time, it can also help them to early detect any chronic problems which used to take time right back in the days. So this will help them based on that all that history right because now AI is trained around that.
At the same time, patient will feel supported and connected they’ll feel connected. And also, this also overall help reduce the stay of the patients in the hospital. If you see that all this remote care or the patient experience and all through AI, they can have everything at home. They’re monitoring different tasks, different automations and all, from different devices. So that the hospitals can be used really for the patients who require them, like who are really in need. And the other patients can go back home and they can be tracked and monitored at their home.
Beau Hamilton (25:48.11)
A lot of benefits there, yeah. I mean, from the provider side of things, it’s like you’re able to actually spend more time with individual patients. You don’t feel like you’re super stressed and you don’t get burnt out. And then from the patient side of things, I mean, there’s a lot of benefits there and not having to even go to the hospital in the first place because you have all these tools available to you at home and able to actually give you the information you need and treat your issues. I think that’s very valuable. So yeah, I think that’s a win-win for everybody.
Now on the operational side of things, one of the main pain points has been dealing with so many different patients and kind of the chaos around healthcare and treating so many different people. What’s AI doing to help on this front to help just streamline things and cut down on errors, specifically in regards to maybe billing? Because we haven’t talked about billing yet, but that’s a big part of the operational challenge.
Deepak Borole (27:03.294)
Yeah, so billing definitely has been a complex part of the healthcare, right? There are different claims which need to be filed by the healthcare organization, right? Which then sometimes they get denials and all. So what AI is doing is basically it’s helping to automate a lot of different steps which happens in the billing cycle, right? Whether it is coding, whether it creating the claims, then checking the data in the claims to reduce the denials. So all this agentic process will help all that out and reduce denials to almost 80-90%. In fact, you won’t believe that lot of health organizations have already started using AI agents for revenue cycle management because it’s admin work end of day. It’s not directly dealing with the patient. So it’s basically when the encounter is done, what hospitals organizations do, the admin don’t do with the payers right? So this definitely has been a high value use case which everybody is using and getting the ROI out of it.
At the same time scheduling and all also has been one of the kind of I would say challenge where there used to be staff who need to answer the phones right? For scheduling, rescheduling, cancellations, or any questions and all. So now what AI is doing is, you might have heard about voice agents, right? Now there are voice agents who are trained around the clinic and the organization’s care and availability and all. So they can schedule appointments. At the same time, they can also see which all providers are available to give a appointment slot to the patients and if they want even if they want to go for any cancellation rescheduling. Everything is happening through smart voice agents nowadays. So it’s very minimal To minimal that they have to reach out to the staffs. And what is what is happening? Is that this voice agent themselves are able to identify? Okay, this is basically a problem which should be answered by a staff member. So they kind of direct the patient to them. So imagine that nine out of ten calls are handled by them and one call might be transferred to the staff. So that’s pretty usable.
Beau Hamilton (29:38.182)
Totally. Yeah, it adds up and frees up some valuable time for other tasks. I just think of my scheduling tasks for my healthcare visits. And when I have to talk to scheduling, whether it’s in person or over the phone, it just seems like they’re very stressed and constantly very busy and constantly playing phone tag with their, trying to figure out who to relay the message to. So any help in this regard, I think, is a win. But I think in general, think all that with the billing and the scheduling and the other kind of upsides with this tech, I think it’s very exciting and kind of paints a more optimistic view of things to come in this space.
But that being said, I know a lot of what this new tech can do. It’s often easier said than done, right? Integration is often, I think, kind of the biggest question mark, especially for teams juggling a bunch of different systems. How do you approach bringing automation into an existing hospital network or a tech stack without disrupting what’s already been working and what’s already been in use for X amount of years?
Deepak Borole (30:47.508)
Correct. As you said, that integration is very, important because there are so many systems in the healthcare space which are being used by the hospital members and the supporting staff. They need to talk to each other. If they work in silos, there is no point. It will just increase more time and effort for them, right? So interoperability is very, very important when it comes to any AI implementation. And that’s what exactly we do. We have 25 years of experience working on different integration in healthcare space. It can be any platform using the HS7 and Firebase standards.
So what we do is we try to implement AI into the platform which the providers or the staff members are using and that integrates into the other system. So that way you don’t have to go out of the platform to do anything. So it is inside the platform. At the same time, it talks to the other platform to get the required data or push any data back to it. That way it works in conjunction and not in silos. But this is a good question to ask that is very, very important for any healthcare systems to have it interoperable. That way it can get the required data for any patients from different systems and use that data to create an outcome what AI is looking for.
Beau Hamilton (32:20.366)
Yeah, because it really just boils down to two decisions, right? Do I build something custom for my unique workflow, my unique needs, and my setup, or do I buy something off the shelf? And I think it’s a daunting task for a lot of providers kind of weighing these two options. But knowing that you work closely with representatives, I think that’s a big plus, and having that kind of consulting sort of workflow with clients, I think will help kind of ease some concerns. And just the fact that you’ve been working for, in business for 25 plus years, I think that’s kind of speaks to how capable you guys are, right?
Now, when it comes to agentic AI, is the real incentive to go custom, especially for like healthcare specific use cases? Like maybe, I know you’ve kind of touched on a lot of these examples, but maybe you can help kind of summarize the real sort of benefits of going with this accustomed solution as opposed to something off the shelf.
Deepak Borole (33:28.348)
So, off the shelf tools are good. I wouldn’t say like, obviously they are good for common use cases or generic use cases, which fits everyone. But they cannot be a good solution for specific health system problems, or where they have specific data. So, if organization looking for a custom workflow implementation or a custom use case automation. And obviously each organization has different data. They should go for a custom implementation with a vendor or a implementer, right? And identify what kind of ROI they’re looking for, right? That’s where the off the shelf versus custom differentiation lies.
And I would say that, off the self, as I said, it’s definitely for quick start, but there can be some limitation if you have more things to do. Custom definitely, it depends what kind of implementation you need. It can also integrate into the different platforms, including the EHRs, right? Or different, it can also include different rules or guardrails which you’re looking for, depending on what data you have. And definitely every organization have a different processes, different staffing models, different compliance needs, right? And that is where there is a need of custom implementation, right? So all in all, I would just summarize that to maintain that control over that workflow, having a secure, interoperable implementation, and also have a long term solution, a custom solution would be beneficial in long run.
Beau Hamilton (35:27.724)
Yeah, I mean, I know healthcare solutions, there’s rarely a one size fits all approach or a solution out there. So like off the shelf solutions, I think they might work in some cases, but there’s gonna be that custom kind of implementation flexibility that is required and then various degrees of that. So I think that’ll be super appealing to know that you’re able to kind of work with whatever sort of system and challenge or platform that’s being used.
You mentioned ROI. So I’m curious, how do you make sure that clients see real ROI and not just in terms of cost savings, but like an actual speed and outcomes, as well as like maybe long-term scalability? Because that’s a big factor of when you’re maybe, I mean, companies are always kind of either growing or shrinking. So you want to keep scalability in mind. How do you factor in ROI?
Deepak Borole (36:19.026)
So basically what we do is we have been using our framework which we spoke earlier where we have defined different steps. So what we do is we work with our clients to identify what is their high value use cases. What are they actually looking for? Are they looking to save time? Are they looking for automation in different processes? And based on that define the ROI right? So they are like immediate ROI which they’re looking for? And there are something indirect, right? So what we do is we define those we define a approach and a blueprint that how that implementation can be done and then come up with that this is what ROI you can achieve based on this implementation.
Because to be realistic we cannot get each and everything, right? We cannot have a solution with like a long list of ROI. So we have to prioritize what exactly you’re looking for and based on that design the solution, show what, okay, what you can get out of it and then go ahead and implement it. At the same time, the benefit about working with us is that we collaborate with our customers. We continuously work with them to show them the progress, the outcome of it that way even if they want to make some changes while they are building the platform we can do it so because ideas can come up while we building something, right? So we can easily easily manage that and that way towards the end they have a product which they’re looking for, right? So that’s what we do.
Beau Hamilton (38:05.454)
Right. So you’ve provided a lot of great general use cases and solutions, but I’m curious, do you have any results or case studies where Chetu has sped up delivery, improved outcomes for clients? Anything specific that comes to mind where you’ve helped a client?
Deepak Borole (38:24.148)
So we have worked with a client where we have implemented agentic AI, a revenue cycle management engine, where what we’ve done is we went through the EDI files, which are used commonly in the revenue cycle management, and that’s where a of denials happens.
So what we did was, we created an agentic task to go through all those claim files, identify any gaps. So before even they are submitted to the payers or clearing houses, agentic AI helps them to correct that data and then submit it. That way the denials are reduced. So you’re almost able to reduce the denials by 80 to 85%. It also helped to kind of revalidate the EDF files up to 70%. And at the same time we also improve the compliance based on the peer requirements. So all that was included in that agentic AI implementation. That was one.
There’s one more where we work with one of our client who were generating some clinical reports manually. So we work with them to create an agentic AI process to automate that completely right? That way whatever the staff they had in fact they were spending 200 manual hours per month so all that was saved through this automation. And they used to take three days to create a report. So that was reduced to one hour with that agentic AI process. So we were able to at the same time not only reduce the time but also improve the accuracy as part of it.
Beau Hamilton (40:21.26)
Yeah, absolutely. I mean, that all adds up saving 200 hours frees you up to do all sorts of other things. And a lot of these things that the, a lot of these AI, you know, tools, agentic AIs is tackling are things that I would, think it’s safe to say that a lot of people don’t like to do in the first place. A lot of it’s just kind of mundane, monotonous work. So there’s that. And I think if you, I think, yeah, if you’re able to save 200 hours also improve, the denial rates by 80, 85% among other things. That’s huge, and it all adds up.
So I appreciate all those insights you shared with us. And I think for prospective customers, just people in the industry interested in working with you or interested in learning more, at least, about some of your solutions, think those examples really kind of resonate with them, I think it’s to say. For anyone listening who wants to explore what Chetu can build for them, what’s the best way to get in touch?
Deepak Borole (41:26.74)
Absolutely, so we are here to help first of all. So anybody who are interested do visit our website. We do have healthcare, a web page they can come in we have free demos over there. We also have sources or channels to get in connected with our members. We can definitely talk to you work with you to understand what your pain points or need is and let you know how we can help you out.
Beau Hamilton (42:00.206)
All right, that’s awesome. I appreciate all the insights you shared with us around this topic of agentic AI. Very exciting. Deepak, appreciate it. Appreciate it. And I hope to have you back one of these days.
Deepak Borole (42:11.08)
Absolutely. Thanks a lot.
Beau Hamilton (42:13.152)
Absolutely. Yeah. Thank you all for listening to the SourceForge Podcast. I’m your host, Beau Hamilton. Make sure to subscribe to stay up to date with all of our upcoming B2B software related podcasts. I will talk to you in the next one.