GlobalEdgeTalk

Navigating Healthcare Innovation with Agentic AI and Human Expertise

• Alex Romanovich

Share episode

Send us a text

The worlds of healthcare and artificial intelligence (AI) are colliding in revolutionary ways. In this thought-provoking conversation, Alex Romanovich speaks with Isaac Chapa, CTO and co-founder of Simpl Healthcare, and Mihir Shah from NetWeb Software about how Agentic AI is transforming digital health platforms.

We dive into the real-world implementation of AI in clinical settings and regulated segments, where Simpl Healthcare has created a guided approach that prioritizes accuracy over speed. As Isaac explains, "We are giving the AI engine specific direction on what to look for and what to output based on the input data." This measured strategy has produced remarkable results, significantly reducing clinician workloads while maintaining data integrity.

The discussion tackles the provocative question: Will AI replace software engineers? Despite bold proclamations from tech giants, both experts emphasize that human expertise remains irreplaceable. The actual value comes from understanding how AI changes the development process rather than eliminating it. "AI should be thought of as tools to optimize, not replace," Isaac notes, comparing the evolution to how spreadsheets transformed accounting without eliminating accountants.

Specifically in healthcare, AI offers solutions to the critical challenges facing the industry. With projections showing significant clinician shortages in the coming decades, AI-powered efficiency tools could help bridge these gaps while improving patient access to medical records. Yet these innovations must navigate strict regulatory environments where patient data security remains paramount.

Whether you're an entrepreneur exploring the implementation of AI, a developer concerned about the future of the field, or a healthcare professional interested in technological advancements, this conversation offers valuable insights into balancing innovation with human expertise. The future belongs not to those who fear technology but to those who can bridge the gap between technical capabilities and real-world business needs.

Listen now to understand how the collaboration between human ingenuity and artificial intelligence is reshaping healthcare delivery for the better.

Support the show

Alex:

Hey, hello. This is Alex Romanovich and welcome to Global Edge Talk. Today is Wednesday, april 16th, and my guests from the technology world and AI world are Isaac Chapa and Mihir Shah. Welcome, gentlemen. Thank you so much, alex. We typically conduct these interviews with a lot of entrepreneurs and, from what I understand, isaac, you're not only a chief information officer in your company, which is Simple Healthcare and Digital Health, but you're also a founder, entrepreneur and CEO of the company. Is that correct?

Isaac:

A CTO. I have a partner that's CEO, so there's two founders in Simple Healthcare. But yes, Paul Jay.

Alex:

Okay, that's great. Let's make a few introductions. Isaac is associated with a company called Simple Healthcare which is in digital health and patient continuity and they built a really amazing modern digital health platform which basically covers the entire HR ecosystem, and here it actually helped build that system. It was involved with a lot of the software development and product development as the lead application development executive with a company called NetWeb Software, which is stationed here in the United States, india, uk and globally Very much aligned and similar to where Gemma is located. I think we're going to have a great discussion. We will talk about agentic AI. We will talk about artificial intelligence and technology, modern ways that software or agents are being built right now, so I'd love to jump right into it. So let's talk about agentic AI. Sounds like a good topic. To begin with, isaac, you've implemented AI inside of the solution, inside of the company, to work with medical records, to generate clinical notes and do so much more. How is it working out?

Isaac:

It's sort of fantastic. It's beyond of what we expected. Always, when you're developing something in a lab in a test environment, keep on playing with fake data, so you never really know how truly it's going to come across. We've had an opportunity to have this in real life, in production, and so far the feedback has been tremendous One. It's accuracy of what we've developed, the speed of which clinicians are able to get their work done. It has been fantastic.

Alex:

Great, wonderful and Mihir. Just a quick question for you Since you've been around for quite some time with application development, building AI solutions, is it any different than building your regular software platform solutions or software products? Than building your regular software platform solutions or software products, and how? What was your experience working side-by-side with Isaac on this particular project?

Mihir:

Thanks, alex. First of all, I'm very grateful to have me on this podcast representing NetWeb Software. It's very pertinent questions. On design engagement solutions so how it is different. On designing agent solutions so how it is different. All customers across the spectrum are asking for the secure, more scalable, very robust, 24x7 available solution where the AI solution should be working very optimised way. We are following that architecture, right processes in designing that solution, especially working with isaac, as they are from healthcare. Patient data security is paramount, so providing that secure compliant there are with all domain, like in healthcare there is a hipaa compliant or soc compliant, so providing those compliant software solutions are very important. So, yeah, speed, accuracy, auditability, performance, scalability are the pillars which we make sure that while designing an AI solution.

Alex:

Now you mentioned something very interesting. You mentioned accuracy. In what way are we able to rely on AI solutions in terms of accuracy? Is it because we're still providing the data we're still providing the patient data, we're still providing the doctor's data, we're still providing whatever other information that we have available or is it because we're now relying on a side-by-side helper called AI to interpret this data differently, to maybe do some online research quickly? Or why is this more accurate? Anybody, please jump in.

Isaac:

Yeah, I wouldn't say it's more accurate at this time. Right, I think it's a tool, and how you use that tool will determine the success and the accuracy of your application. So, at a high level, I would not say that providing AI as the standard input of medical records it's going to diagnose cancer or anything like that, right, it's simply going to go use its historical knowledge of data the internet essentially and provide a mathematical algorithm answer based on that, and that will vary. So we at Simple purposely do not try to diagnose or do things of that nature, and I think that's the difference between AI. And then the topic of your agentic AI.

Isaac:

Right, with agentic AI, what we've done at Semple is two things that are a little bit different. One is the input data is data that we are creating, right? So, whether it's the medical records, we have custom patient forms, patient note forms. These are data inputs that we are collecting from the clinicians and the patients. The second thing is, now that we review that input, we're not asking generic AI to summarize it. We are giving the AI engine the direction of specifically what to look for and what to output based on that input data. So in this case, we're guiding the accuracy by basically telling it if you see X, y and Z, give me this output. If you're seeing A, b and Z, do this instead. So we're using the workflows of AI, but we're not trusting AI at this point to give us. We're not basically saying go do X, y and Z for us. We're telling it that, with this input, produce this output, and that helps us get a much higher accuracy as opposed to just a generic LLM.

Alex:

Was that your choice? Because a lot of consumers, for example, are using AI now as their clones and even before they visit the doctor I know I do it before they visit the doctor, they bring this entire output that says I took a picture of my pimple, or I took a picture of my tooth, or whatever, and this is what AI told me. This is what JetGPT told me. Are you being extra careful about this, or is some of this data also making its way into the LLM or into how you interpret this information?

Isaac:

Yeah, I think we want to be extra careful with it because, again one, it's medical data, right, so we need it to be as accurate as possible.

Isaac:

So it depends on the use case. If a healthcare organization comes to us with a very generic use case that says, I just want an overall summary of the information, Okay, great, that's not something that is imperative to their operations, it's not something that is going to misdiagnose, that can use a generic AI solution. But if it's something that they are producing their clinician notes, it's something that is producing their ICD codes, their workflows. We took the decision of accuracy over speed and off-the-shelf solutions. So that's where working with Mahir and team have optimized that, and I think, just to say we've worked with Mahir for over 20 years now. So I think just to say we've worked with Mahir for over 20 years now. So I think our working relationship goes back to previous, to Simple, to FinTech and other organizations, and so we have a working relationship where they provide a fantastic foundation of resources and tools and we come with business logic and workflows that then take those tools and essentially bring it to another level.

Alex:

Wow, you guys have worked together for 20 years. You started working together since 10 years of age or something.

Isaac:

Yeah, we were both 12 years old when we started.

Alex:

What do the same kindergarteners Question for me here. So, over the course of the 20 years you being an application developer, integrator, software developer 20 years ago we were still especially in healthcare we're still dealing with the paper forms, dealing with faxes by golly, we're dealing with this right now. What have you seen in terms of software development advances over the course of 20 years and probably we could easily say, over the course of the last five years? Give us a little bit of a review of what you've done five years ago, 10 years ago and what you can be doing right now as a software developer.

Mihir:

It's a very interesting question, alex. Thanks for asking. Yeah, so if I say from last five years we were doing application modernization, every customer, every enterprises, are going moving to cloud. They want scalability, reliability, cloud migrations and now, with every use case, they want more and more automation where we can save their overall cost of development. Now they want product ai, can embrace ai in what they are doing. So since last three, four years at NetWeb, we are continuously working with all customers where we can embrace AI more and more, which can efficiently add more value towards their end customers. So that is the big shift and I'm seeing myself be developing an agent, be development, testing, designing UI works, how efficiently we can do it and how we can bring AI to help the real world. That's the leadership that we are seeing and, with the ISA and Simple Healthcare where we are working, that shift has helped our association and our collaboration.

Alex:

So first, microsoft chairman and the CEO says that software platforms are dead. It's going to be the world of agentic AI. Now, google recently comes out with a new framework based on Gemini, but it's a different kind of Gemini that basically says look before, we're not even going to need the software engineers. It's all going to be integrated together, different agents will be working together, and so forth. Are we ever going to see the time where we're not going to need a software engineer? Is that time coming, or is that little too science fiction and premature?

Mihir:

As per me, like personally. Definitely it is going to impact the software engineering practices, but going to amplify the overall practices. The more we will address AI, the more you address AI. Whatever the output these tools are generating, for an example, like autopilot or beat cursor, the agents whatever the output they are this AI tools and services are generating, you need to have that human oversight that, whether those outputs are correct or not, whatever the code they are generating, you will definitely need senior developers to make that more scalable, more secure. Okay, as an integrator, we know, like how progressively those functional pieces can be integrated to make that entire into an enterprise solution. Work will be become a guide or a mentor to use, like what type of AI tools, what type of agents needs to be built, what type of frameworks needs to be built, what type of design patterns needs to be built, how the deployment should be done more efficiently. So definitely the years will change, but the overall engineering practices will not affect it. That's my opinion.

Alex:

Isaac, what are your?

Isaac:

thoughts on that. Yeah, I agree. To take an analogy, right, think about accounting years ago, prior to spreadsheets, right? You had accountants who were using paper ledgers to write their work. Then it evolved the spreadsheet came along and they started using computers to do that.

Isaac:

In general, ai should be thought of as tools to optimize, not replace. At this time, I want my search engine to give me accurate AI results before I start thinking that it can code our programs. Right, because at this point I see a 50% accuracy on some of the things I have the search engines in regards to its AI answer and I end up skipping it at this point. I wish there was a feature that lets me skip the AI output because it's been wrong half the time. So I think, in general, your organizations have to use these as tools. So you know again, me developing software for over 20 years.

Isaac:

We've seen different types of developers. There's developers who go out to the internet and search for code and then copy and paste the code and then plug it into the environment and it causes problems. Right? That's a junior developer who doesn't have the experience, and they're basically in chat rooms or looking for Google on how to answer their solution. So, in general, ai can do the same thing right. So, to Mih mihir's point, the entry-level development tasks could potentially be automated through ai. However, the problem with that, as I mentioned, is those junior developers have introduced bugs because they don't understand the business logic or the deployment process.

Isaac:

You're going to have the exact same issues with AI, where a senior developer, an architect, someone, still has to interface and understand that the applications impact to other applications. Did it meet the business requirements? Sure, functionally it's what it's supposed to do, but will it have any impacts? And in general, that's still always going to be there. There's no way that AI has the knowledge to solve that problem today.

Isaac:

And if you think about it, where has AI been trained? From coding the Internet? How many private companies' source code lives on the Internet? Very little actual content on the internet for code is help support open source, right. So I don't think it's what I would call the best quality data to model your ai on. So if you're using the internet, which is primarily troubleshooting now, ai is going to write code. That is basically when with bugs that it's looking for help on. So I think there's still some time to either build a model that is specifically to your data set, your code that you can optimize, but it still takes some human tweaking to make that available and what I would call enterprise ready.

Alex:

Yeah, and would you say that AI maybe is better suited for? We'll always need junior developers, right At some point in time. I used to be a developer a long time ago and I had to start someplace, right. You, gentlemen, you did the same thing, right. So we're always going to have junior developers that are probably going to be doing things maybe a little differently, right, and instead of copying, pasting code, they will be using AI to do some simple tasks or things of that nature. But can AI, within the context of software development job for AI, for example, to go and clean some of the things out there, maybe uncover some missing links or uncover some missing relationships, or something like that or could it even sweep after the junior engineer and make sure that that is also sound? So AI may have some kind of a maintenance or QA, quality assurance type of a job, if you will, within the being a member of the team or being a part of a team. What do you guys think about that?

Mihir:

I remember me and Isaac were working on large tech software where we used to provide such cleanup jobs to junior developers and after that you will see some bigger issues coming from the production side. Okay, so definitely ai can do that job where ai agent can clean up the code. But after integration testing, overall uip testing, that code is not affecting the corner case scenario hk, still human oversight is required, but definitely it is going to speed up that process. So whenever you are designing any agents, whatever the core output it is generating, you will require a different agent when you're making that output. If there are any critical steps to be taken by that agent, then definitely human oversight is needed. Yeah, that's my idea.

Isaac:

Meher said a big word speed. That's what every business person I've ever worked with wants from the dev teams is fast, right. That's the number one thing is how fast can you do this as opposed to how good can you do this, so it's always a combination of that, right. So I think using these tools from a speed perspective is very interesting, because you talked about the cleanup of code. I've met so many developers who are frustrated with just how the code looks, and their first task before they even start troubleshooting your issue is let me clean the code up to where. Then I can diagnose what I need to do with it. So you could use AI to ideally copy and paste this, but make it cleaner, make it easier to diagnose, right. So that can definitely help with speed. Instead of a human spending a couple of days cleaning up and making it look pretty, feed it to an engine that will optimize the look of it and not change the code itself, right. So that's a big thing.

Isaac:

And then two, I would say, the troubleshooting aspect. That's a hard job. There's developers who love just to troubleshoot. They'd rather troubleshoot as opposed to develop new code. So it's a very specific skill set and so it also takes a lot of time. Right, there's so many bugs we've worked through for the years. Right, I don't want to say I'm a buggy code all the time, but you always have these critical issues or major issues, and people teams will spend hours researching that data. What we found is that you can actually potentially produce a Kubernetes log into AI and AI will give you an hint or potentially the answer. So there is a value of using the error codes, the diagnostic tools, with AI to say, hey, here's what AI thinks the source is. That can help speed up the diagnosis problem. So I think both of those really are around speeding up the process, getting quality improved, et cetera. Those are two good use cases for AI.

Alex:

Excellent. Now, you guys are very unique in the sense that you work with both in financial services and then healthcare to highly regulate, and what's interesting about that is that I'm not sure how mature the legislature is in terms of regulation of anything that's AI-based, ai-developed, ai-integrated, ai-verified and AI on and on. What are your thoughts on artificial intelligence and how it's being developed and how it's being deployed in large organizations, healthcare organizations, smaller companies like Simple Healthcare, in highly regulated industries? How's AI being treated and almost regulated?

Isaac:

Yeah, so different points here, so I would say let's break it down first. Large organizations, small organizations I would say the commonality is everyone is trying to create efficiency. One of the biggest challenges with healthcare is, you know, burnout and lack of resources and lack of clinicians coming out of the market, right? So in a couple of decades, there's going to be a major issue with not enough clinicians and a lot of sick people, right? So I think everyone has the same mission is how can we use AI, data and technology to improve those areas and do that? Right? So I think, fundamentally, people all have the same goal and let's figure out how to solve that. The larger organizations they have a lot of funding, so I think they will take a much more pragmatic approach to it to say, let's trial and error and see where we can make those improvements.

Isaac:

Small organizations I would say break it into startups versus healthcare organizations. With startups, we have the mindset of again being from a fintech. It was protecting data, gdpr, ccpa, et cetera. We had that ingrained in our head around data security, people's information. We were a part of this company. We did identity theft and fraud protection. We helped some of the largest data breaches in the world over the last 15 years. So we firsthand how data is misused, et cetera. So, from our perspective, we understand it and we go and treat it with as much respect as possible when using those tools.

Isaac:

I cannot say that every startup is going to do that, right, a lot of startups are going to do whatever they need to do to make money. So, you know, a lot of those people may not protect the data, they may not do the right things with it. And there's your standard legislation, right, your HIPAAs et cetera, that provide protection against the data. But until they either lose or mishandle the data, somebody catches them. There's not some way to continually monitor and regulate that and I think, our regulations from a government perspective, especially here in the US.

Isaac:

I don't think you have a lot of security experts writing legislation, right, a lot of the legislation comes from people that are influenced by different organizations on how to write that legislature. So I think there needs to be more of a let's get together a group of experts that can help advise the legislation to be written, and I think some of those, like with the Cures Act and interoperability and data sharing, they've got to about 80% there. Those last 20% they're not in the weeds, understanding the use cases and the data. So it's still a gray area of how can the data be used or how can we get this to benefit both patients and providers. I think there needs to be more experts in the room, but that's very difficult because the government's writing regulations on a bunch of different topics so it's very hard for them to go after experts in every single piece of legislation that was being written.

Alex:

Another question I have is we probably had more innovations in the last five years than we've had in the last 20, 25 years, right? What do you personally, Isaac, expect of tech innovations as you take the company to the next level, add more functionality, enhance your roadmap and so forth? What can we expect from you, or what can we expect from the industry in general, Provided that it's regulated, there may be some slowdowns, but can we expect some dramatic breakthrough over the next three to five years?

Isaac:

With all technology, you get major leaps and bounds from year to year. If you think about 20 years ago, cell phone technology and what they look like to today, it's a dramatic difference. So you're always going to evolve and get new technologies, I think. For Simple, specifically, our focus has been two things right, I mentioned earlier the clinician operations, burnout, et, etc. We're going to continue to focus on how we can improve the lives of clinicians and healthcare organizations to be more optimal, a lot of tasks that are currently human-driven. We're working on taking that and automating those functionalities. So our roadmap is specifically around how do we help the clinician and healthcare organizations.

Isaac:

I think, secondary to that, our focus has really wanted to look at the patient access to their medical records, and here in the States it's very rare that you have access to your medical records or have ever seen anything as a whole, and so, with the Cures Act and some of the legislation coming out, we think over the next couple of years that's going to start being more prevalent. So, just like you get a credit report today, you could potentially pull your medical records, and so I think that's very important from the perspective of one the patient has access to the data, but two they could share that with their oncologist or other areas that may need that information, because that's been difficult to get. So I would say again, our focus is not providing the care, but we can enable a way for data sharing to get patients better care by giving them either the provider or the clinicians access to that. So for us, it's really around solving those pieces. Right now, healthcare has so many issues as it is.

Isaac:

We can't solve everything, but I think fundamentally focusing on those two areas for the next couple of years will help us thrive as a company and with our partners. Our model is around B2B, so we help other organizations with their needs. One or two or many of the companies we work with are trying to dive deeper into solving certain health issues or looking at data. So if we can be a part of it, that's fantastic. Right, we could partner with a company that or a healthcare organization who comes up with a certain medication for something based on the access to this information. So I think that's what you'll see is, once you have more access to data, there are now more products, more solutions, more innovations around that.

Alex:

Hopefully, ai will be that type of a catalyst, will be that type of a glue that will bring it all together for the consumer. We have to make things simple, but not necessarily simpler. Those are not my words, those are the words of Einstein. But what are you going to be looking for as you go forward as a software engineering, software development organization, in terms of what type of skills you're going to be looking for? What is going to be the team composition, if you bring it on board and growing new resources, new engineers, what are they going to look like?

Mihir:

In software development right now, like we are for any team composition, we get along with the customer. So if we become as a technology partner so for an example, with Simple Healthcare with Isaac he brings sort of great vision. Always we work along with him in sharing that vision. That collaboration is the key to success. What is their vision? And understanding their vision patiently, chatting out that team members like what type of UI UX will look like, how the architecture will look like, how we can align more Isaac's and his team's vision, is very important. So deciding the technology framework, deciding design frameworks, deployment strategy, testing strategies Accordingly, we make a team.

Mihir:

What we look for people is your their good analytical skills, whether they are good problem solvers or not. Continuously in AI and agentic AI world, there are continuously new elements, new tools are coming up. So are they good learners or not? That's what we see. The culture of NetWeb is comfort-made. We are precision makers we call it in medical terminology. We listen to magnitude. We are precision makers we call it in medical terminology. We listen to patients. We are the surgeons. We accordingly frame the team.

Alex:

Interestingly, you mentioned the human side of development, the human side of working with software development companies internally, externally. That is still going to be important until robots begin to work with robots. But I think the human component is still going to be important until robots begin to work with robots. But I think the human component here is going to be extremely important In this very cold world of automation, innovation, tech. Humans are still going to be important in collaboration, in communications, obviously, and is that the same true for you, isaac? Is that what you're going to be looking for?

Isaac:

Yeah, I think there's a misunderstanding in the software development lifecycle that if a project takes, let's say, four weeks, then you have developers coding for four weeks, and so there's this understanding that oh, it takes four weeks, it's going to take four days. Instead, the coding portion of it is not four weeks. I think that the coding portion of it could be, right now, maybe one week. You're spending three weeks of requirements and design and iteration. And that is the issue right now, with folks who are not familiar with software development thinking that ai is going to solve everything.

Isaac:

It's not solving requirements and design, the I does not know. Do I need a new microservice for this and how do I want to split it out? What are the effects of the other services? And that has always been frustrating. It doesn't matter whether it's today or if it was 20 years ago. A lot of business folks without any background in software development just here. Four weeks is too long, when three weeks of that is just the requirements and design phase of it. So you need those humans to really work together and I've seen a lot of instances where the humans themselves do not work together well between product and engineering, right? So there's so many companies, large or small, right now, that still have inefficiencies, and it has nothing to do with the actual development itself. It's the requirements, it's the design, it's the part to get it ready for the development code.

Alex:

So I think it's extremely important by the technologists in the organization, by the marketers in the organization, by the business unit executives, to really come back to the CEO or the C-suite and say look, just because it's AI, it doesn't mean it's going to be necessarily faster or initially cheaper, and that's something that CEOs don't want to hear. Obviously, they see a big shining object called AI. They want to implement it yesterday and they want to deploy it yesterday and they want to gain competitive advantage yesterday and so forth. It's been a fascinating discussion. As a parting word from both of you, I would love to hear your advice to the upcoming or existing entrepreneurs who are building products right now, whether regulated or deregulated industries, and also a piece of advice to some of the younger software engineers and people who want to get into the field and scratching their heads and saying, if AI is going to develop everything for me, maybe that's not the right field for me to go into. Gentlemen, in your own words, what should be the advice to those respective individuals?

Mihir:

Yeah, for the enterprises, for founders, startups, customers definitely AI is going to rule. The more you embrace AI, the more you will need the life of your end customer easy. For an example, in healthcare the current with Simple Healthcare, the AI use cases which we are creating it is definitely helping the end customers, the clinicians, the patients. The new use cases where we have created agents which transforms the core, where we have created agents which transforms the core written in older technology to newer technology, definitely saves your 15-20% of over time and effort. So the more you embrace AI is going to help.

Mihir:

But how to use it, what should be the AI governance framework? What would be the security framework? How the design should be? Those would be the interesting points and where the companies like NetWear will definitely help. Speed and there are continuously newer changes keeps on coming. Google has come up with now agent-to-agent protocol where you develop by agent and agent will communicate with other agent on that protocol. So every day there is a new news. So every day there is your news, every day there is your innovation. So to embrace with that innovation is the key in the coming times. Isaac, your thoughts.

Isaac:

Yeah. So there's two targets, right, you have the entrepreneur and then the developer. So, on the entrepreneur side, what I would say first of all is it's difficult, right, this is not an easy task to go commercialize some idea into a company. So, first and foremost, be ready. Right, this is not going to just be an easy task. You'll have a lot of challenges in regards to the economy. What's going on in the economy? Things change in regards to the economy. What's going on in the economy? Things change in regards to funding. You know, I think during COVID, if anybody said they had a healthcare idea, they were getting funded. Now, five years later, you need to show revenue just for an idea to get funded, right. So you're going to have ups and downs based on what's going on in the world economy and everything else. Buckle up and be ready for a lot of ups and downs. One day will be exciting and great because you got your first contract. The next day is going to be, you know, disappointing because you lost three other ones, right? So, from that perspective, just keep your head down, work hard and just grind through those phases of your company and I would say, being through the phases of five employees, to an exit. That doesn't change, it just happens at a greater scale, right? So every couple of years you're a different inflection point. So you're constantly going through that From the developer perspective.

Isaac:

The funny thing is, you know, back in college in the 90s we were told the same thing. We were told don't get into computer science, it's going to be automated, there's no value in you getting a computer science degree. So I've heard that for 30 years now and what we've seen, I think, is I've helped advise for university locally. There's still a challenge with them trying to fill the seats in the programs. So people are hearing that message saying oh, I have been told don't do computer science, and they've heard that for 30 years.

Isaac:

But what I think is missing is the piece saying look, we need humans who understand technology and ideally business as well, that can work with the, that can work with the AI, that can work with developers, that can essentially translate.

Isaac:

And that was essentially my skill set coming out of college was I could take technology and work with developers, but I could sit with a business team as well and understand business and essentially translate it. There were so many times over the years where the business team and the tech team were saying the same thing, but they were arguing because they didn't understand each other, and so I think that has to be. The focus of technologists coming into universities is how can I help fill the gaps? Because if you can fill a gap, then now you have a rarity and you can make more money. You can get more opportunities because you're solving problems that have not been solved yet for over 30 years, and if somebody just wants to go sit in a typical encode great, you can get your college degree also and start coding. But I would say filling those gaps is something that is not focused on when it comes to universities and teaching computer science resources.

Alex:

This resonates so well. Just to be a good prompt engineer, you need to combine those skills right the technology skills and the business skills and even to define the problem, to define the technical goal and objective, define the architecture and so forth, you need to communicate. Gentlemen, it's been a pleasure, extremely insightful. I think our listeners are going to be very excited to learn more or validate a lot of things that they might be thinking about as well, both on the business side, tech side, innovation side and product development side. I want to thank you for being our guests and I hope we can continue to have this conversation. Thank you so much and I wish you all the best luck in your endeavors.

Mihir:

Thank you, thank you so much, thank you so much and wish you all the best luck in your endeavors. Thank you, thank you so much. Thank you so much, alex Isaac, for your time and for the opportunity. No-transcript.