Massoud Alibakhsh Podcast Transcript

799
Headshot of Chief Executive Officer Massoud Alibakhsh

Massoud Alibakhsh Podcast Transcript

Massoud Alibakhsh joins host Brian Thomas on The Digital Executive Podcast.

Welcome to Coruzant Technologies, home of the Digital Executive Podcast.

[00:00:12] Brian Thomas: Welcome to The Digital Executive. Today’s guest is Masoud Alibakhsh. Masoud Alibakhsh is a tech entrepreneur with a background in electrical engineering and computer science. He’s currently leading the task of creating the next revolution in human and machine communication at OMEDEUS, formerly Xeba Technologies.

From 2000 to 2018, Masoud was founder, CEO, and Chief Architect for one of the first cloud based medical practice management and electronic health record systems marketed under the NewMD brand. NewMD was a small offshoot startup, which survived the DOT com era and grew from a handful of talents to over 250 employees when it was acquired in 2018.

Today is a part of the Global Payment Network, a Fortune 500 company. Creating multiple software technology startups and building comprehensive operations consisting of production, marketing, sales, support, implementation, and administration has given him a true and unique insight into the fundamental challenges and the urgent need for optimal and synchronized communication, especially in tech producing organizations.

Well, good afternoon, Massoud, welcome to the show!

[00:01:18] Massoud Alibakhsh: Good afternoon, Brian. Thank you for having me.

[00:01:21] Brian Thomas: Absolutely. This is so fun. my audience gets tired of me saying this, but I really just love meeting new people every day on this podcast, but what’s cool is they get to meet the person after the fact after we record and publish, obviously, but this is so, so fun.

And we’re going to dive into some great questions here, which I’m going to start right now. You’ve got quite the career as an engineer. You were in tech you’re entrepreneur, and now the CEO of Omedeus. Yeah. Could you share with our audience the secret to your career growth and what inspires you?

[00:01:49] Massoud Alibakhsh: I think the secret has been my childhood curiosity. That’s been driving me as long as I remember. I was always enamored with automation and technology as a young And that pretty much laid out my path in terms of understanding how to put together these machines and how they tick. I always wanted to build robots and I was always fascinated since I saw my first robot on TV on the, on a show called Lost in Space.

And a lot of your audiences probably don’t know of the shows back in the I think the seventies it was black and white, it was, and they had a robot that walked around and talk to people. And I guess that drove my curiosity, and I ended up majoring in electrical engineering, then computer science, and I got into working with industrial robots, and then later on building financial systems and went back to school to try to understand the stuff that I was doing better.

I got into parallel processing and distributive systems for a while. And then I’d started a company. I always had this knack for doing things on my own, is the entrepreneurial fever, if you will, and building things and creating things. So that’s really my motivation in, deep down, I’m an engineer.

I like building machines and seeing how that impacts the lives of people. And this is my third technology startup that I’ve started. And it’s the most exciting one.

[00:03:13] Brian Thomas: I love your story, Masoud. We’ve got so many people that were influenced by, sci fi, television, books, movies and that’s awesome because we are a technology platform and we talk about tech all the time and I appreciate your story. That’s so awesome.

Masoud, we’re going to switch gears a little bit from you into what you’re doing now. As far as technology, talk to us about your patent pending AI powered software platform. I believe is truly cutting-edge using object messaging and intelligent objects, bringing your project management, your collaboration, documentation, communications all into one seamless tool and how this will bring about better productivity in organizations. Maybe you can do that a little bit.

[00:03:52] Massoud Alibakhsh: Yeah, for that, I’d have to go back and quickly review the history of evolution of software. And I think it would be interesting for your young audiences and probably reminiscing for some of your older audiences because they lived through that process.

Essentially, I got involved in this. As an entrepreneur and as an engineer, when I was that done a few startups and, when you start a company and my last one, actually, I started from my bedroom, literally, and I had an idea about developing distributed systems over the internet, during the DOT com days. And you start from by yourself and then you add another person and then you ended up going to be in 200 plus by the time the company was sold. But as you go through this, you as an engineer, you’re, you realize as you add people to the mix things start breaking down.

So I was enamored with what’s happening, why, as you add more people you become less efficient. And I realized that actually has something to do with communication. And we, just like any software company, we ended up standardizing on JIRA to manage the development. And we had Asana, we had Trello, we had, Slack and for managing communication and email and wrote software to bind all these things together. And we had a document management system and things really didn’t work perfectly. Things were always breaking down one way or another. And that actually really became the incentive for me to pursue this discovery or the inventions that We’ve come up with for our latest product, and I think a lot of people this resonates with the story I’m telling in the way they’ve cobbled up all these tools to try to manage manufacturing software, which is the most sophisticated piece of technology that humans can build, especially in the Internet era.

But if you go back to. The way we started automating software or creating automation with software, we went back to, and it goes back to the 1960s where software engineers went to these companies or corporations and ask the managers about their processes that they wanted to automate. And we would be shown what the process is and how they have these paper forms that they capture information on.

And they. Give it from one person to another. So basically, based on some workflow process that they already had, they captured information on these physical papers and hand it from one station to the next, this is before the advent of computers. And we as software developers, we took those papers and we stuck them on the screens and that model, and we took the data from those screens and put them in a database and did some interesting stuff.

And we improved the standard of communication because software is essentially a communication tool. And we improved communication. We were able to enforce certain rules in terms of clean data capture. We enforced editing criteria, but the model stayed but the same, that form automation and these physical forms ended up being on the screen.

And fast forward, we get to the 1980s. We ran into a big revolution and the big revolution in software engineering was really the advent of the graphical user interface. Or GUI for short and that was a new model that was introduced to the world by Apple and Steve Jobs and their Mac and then followed by Microsoft’s Windows, which made computers less intimidating, and it brought the, in those days, a lot of people may remember.

Computers were not as prevalent and PCs had just finding their ways into society and firstly in the business world and then the homes. Nowadays, everybody runs around with a massive computer in their pocket, which they refer to as mobile phones. But back then in the 80s, when the graphical user interfaces came in, there was a massive revolution.

People may not remember, but the model of building software has completely changed. We went from straight line procedural, take an input, do a function and print an output to stick these objects on the screen and then try to program their behavior and that became the programming model, which is the architecture and that was referred to as event driven programming, which improved.

Human machine interface in a amazing way. And then, and to this day, we’re still benefiting from that. And that really caused a lot of companies that didn’t adapt to this new model to go out of business. A lot of new companies showed up, be the young people, a lot of startups who got excited about this new model.

Cause it was exciting. They’re beautiful screens, but. Having gone through the GUI the revolution, the model, the form model automation still remained the same. In fact we think that it actually reinforced that model because you could really represent the physical forms with higher fidelity on these screens with high res and logos and things that resembled physical entities in the real world, like a trash can or a file folder.

So those were brand new. They were exciting things. If you could drag a file and drop it into. file folder would excite a lot of people in the past, you’d have to type commands, these cryptic commands to be able to move something from some area to another. The replication of the physical processes on the computer stayed the same.

Although the model of programming changed. But that’s because graphical user interfaces were a platform. So, we needed to go through a transition, brand new model of programming to this. Fast forward. We did another, we had another revolution during the dot com days where local area network type.

Computing changed to what we call today the cloud, where we move from that form of computing to distributive computing, which changed not only the model of programming, but it changed the way we were organized in the past, you and I, Brian could sit in our office and the two of us could write a program and put them on this disk or CDs or floppy disk and distribute it.

Today, we need armies of back-end people, front end developers the designers and DevOps different teams with different functions, building a completely different architecture with microservices in a variety of different forms. That’s a whole new mode of computing and writing software that completely changed the way we’ve manufactured software.

Same thing again, a lot of companies failed to adapt and adopt the new model. They went out of business and a lot of new ones popped up. In fact, most of the companies that we know of as software companies, technology companies today, were born during those era. And Google’s of the world, the Facebook of the world.

And Microsoft was one of the. Those companies that adopted this form and embrace the internet in that form of computation, because they saw the turn, but still the form based automation still remain the same. Stick a form on the screen, put the data in there in a structured fashion, age. 56 size 22 and save the screen and communicate it to the next person because the data goes in the database and somehow it shows up on somebody else’s screen.

Now, the reason why that’s important is because that’s a structured form of communication in our processes that are important because we’re manufacturing things, we need exact measurements and exact data. But humans communicate in this fashion, in the fashion that I’m talking to you with natural language.

And natural language is capable of carrying a lot of subtleties of information. And so, email showed up in the corporate world to cover that gap. The gap that the structured communication of the traditional systems could not carry. That email fell short, it promised to be a panacea, but it turned out to be a Pandora’s box.

I won’t get into the model. Why? Because it’s flawed and email is not scalable. And then we fast forward. We started the, what popped up next in the corporate world was what we call the social media networks these Facebooks of the world that showed up in the form of Slack and Yammer that tried to cover the problems of, email and they also promised to solve some problems, but they ended up really channeling information. So we ended up with information scattered in email, scattered in the silos of these channels and scattered in our databases. Fast forward today. Okay, now what do we have? We have AI, artificial intelligence in the form of Large Language Models.

They’ve blown up in the form of ChatGPT. I’m sure all of your audiences have heard of ChatGPT or have used it. And that is really the promise AI come to fruition in terms of number one, the way they process natural language. In a very reliable way, the generative stuff is not as reliable, but we already are using tools and creating a lot of productivity in what we refer to as low hanging fruit because we are applying these tools as tools just one time, load a document in it, summarize it, ask it questions fake somebody’s voice generate videos or all kinds of interesting methods of using this as tool. But our view is that AI is a platform. Just like GUI was a platform. Just like Internet was a platform. We need a new computational approach in building software that allows the deep integration of AI into every aspect of our traditional way of building software.

So that way we can make software intelligent, and that way we can actually truly deliver the promise of AI, all these speculations about how AI is going to change the world and help the world. Our view is that it’s not enough. The LLMs are not enough on their own. You can have ChatGPT 5.0, 6.0, 7.0, it’s not going to be able to handle the sophisticated workflow of these complex organizations with complex workflows, with expertise of people with master’s degree and PhDs in finance, in, in medical technology or doctors or physicians.

And you can imagine all these professionals. In short, that’s the face of the problem.

[00:13:25] Brian Thomas: Thank you. I appreciate you breaking that down and like you, I’ve traversed this career quite a few years in technology and we’ve seen a lot come around. We do a leapfrog as far as tech goes, but you’re right.

There’s things that they haven’t solved. Or for example, ChatGPT, conversational AI can’t solve. And so what you’ve done is really put a solution out there that is going to help us move forward with our efficiencies and productivity. So I appreciate you sharing my next question. I’m just going to jump in here.

And really last question of the evening is you’re obviously leveraging some of that new and emerging technologies in your tech stack. Do you mind sharing 1 or 2 items with us today, as long as it’s obviously not confidential or proprietary?

[00:14:05] Massoud Alibakhsh: No, sure. No, in fact, I welcome the question because we’re holding webinars to try to teach this model to all the young entrepreneurs or any company that’s interested in adopting this technology, even though we’ve got the.

Patents lots of patent spending on this technology. We’re open to licensing it even for free for the qualified companies to adopt this technology. What we’ve done is and I’ll describe how technology works. Essentially, we can’t see as engineers and software engineers, how a large language model.

Is going to absorb a workflow or any workflow, and not anytime soon. In fact, I can’t see it even in the future versions. What we need, what we see is a merger of the traditional model of building software with an enhancement that we’ve added to, which is what we call the object messaging and intelligent objects, and that the model goes according to the following form.

But traditionally, we did the analysis of the workflow. We created workflow diagrams and control flow diagrams and try to mimic that on the machine with forms. And what we’re saying now is we need to go one step further. We need to do semantic analysis as software engineers and identify all the real and virtual objects.

And what we do is when we identify these objects, we need to make these objects self-aware in terms of their object. understanding what they are. And if it’s an x ray, for example, in a medical record, that object needs to know that it’s an x-ray. So that’s structured data as traditional programming.

You basically, let’s say Brian takes that x ray or Masoud is the technician, is a chest x ray. So those are all the attributes of that object. And it’s all structured data when it was taken and it was a chest x ray. And that x ray is aware we’re using that structured data that what it is. And we teach it its workflow that what it’s supposed to do.

Those are all traditional programming that can be embedded in a standard way, and these workflows could be dynamic or static. But what’s added to this, in this object, we basically give it access to an LLM. And this LLM allows this object to understand natural language. And we also provided with channels of communication for each class of stakeholders that this object needs to interact with, we give it a specific channel and that channel is used to communicate in natural language with those stakeholders and not only using natural language to communicate with those stakeholders, but the object uses that channel to move around through the workflow and convert the critical events that are relevant into natural language and place them into those channels and immediately.

Transfer itself to the right stakeholders and communicate those events in natural language to the stakeholders and allow them to interact and respond. And at the same time, those channels are used to load up any information and text file video references. So the object becomes responsible for containing all the information related to the object itself.

I can give you a reference. The inspiration of this model comes from biology. In the back of our brain, there’s not a file folder that keeps track of information about my liver cell. The information about my liver cell is contained within my liver cell. That liver cell is intelligent. It’s self-aware that it is a liver cell in the sense that it functions like a liver cell.

It sends messages to its neighbors. It knows its proximity and it exchanges that proper information with its neighbors and it collects information about itself and the collection of these liver cells create a super object called the liver. And the liver as a super object is also self-aware and it knows that it’s connected to the rest of the system using veins and arteries and neurons.

And these systems are connected with all these super objects, including another object called your brain, which is a collection of the neurons and that super object acts as a conductor of the entire system. So the symphony of all these together in a very intelligent way produces our behavior or a bird’s behavior.

So if you take that model and try to apply that, for example, to an electronic medical record. Imagine when I take an x-ray as a technician and take an x-ray of a chest and that x-ray wakes up knows, okay based on the rules that I have, I can go to the site that’s authorized in private of Mayo Clinic and they’ve got a great engine that can process x-rays and is a great AI model.

I push myself through this. It generates a report and I read it because I can read text because I have an LLM. So I read it and I go, Oh God. It looks like I’m cancerous. So, I gotta run over to Dr. Brian, and show up and say, look hey, I’m cancerous here. You need to take a look at me. This is the object that’s self-aware, and it’s the message.

The object itself is the message, and it shows itself in the right inbox. Of the right priority for you as a doctor and it knows that okay Well, you’re a primary doctor and this is this requires a particular expertise. So by the way, Dr. Brian, I took the trouble of contacting the appointment object of this doctor who’s a specialist in South Africa and is part of our network and we negotiated a Conference call between him and you in two minutes and i’m going to connect you guys in two minutes and boom You’re connected and I’m sitting there listening.

Now there’s a video going on and I, as an x ray, I’m sitting there listening to your dialogue and I’m remembering everything that you’re saying about me. And I’m going to memorize that. I’m I’ve got a memory. I can process that. I can process natural language. And by the way, I’m going to record all this video and I’m going to keep the record myself.

So if you ever need it, I can play the whole thing for you, or you can query me and ask questions later on. If you don’t remember about what did Dr. Jones say about you? So I can actually sit there and record this. And once it’s done afterwards you can talk to me and say, well, I guess we’re going to have to talk to the patient.

So yeah, sure. I just quickly talked to the appointment object and send a message to the patient, having them come here tomorrow at earliest. And by the way, at eight o’clock, you had another appointment that was just a normal appointment. So, I took the liberty of rescheduling that appointment to another hour.

And I’ve already organized everything with all the right objects. And the patient is going to come in. And by the way, I just read the latest HIPAA rules that were published yesterday, and I’m aware, and I can tell you that this is the kind of stuff that we need to talk about, or these are the latest issues that we may need to address.

So we can build systems that use LLMs and different forms of AI engines for particular purposes to make these objects. intelligent for their tasks and the collection of these objects create super objects and these super objects organize themselves into the system and the system becomes responsive to humans using natural language and these objects know how to navigate the workflow they know how long they have to stay in each workflow depending on what the system is so that becomes the creativity of software engineers now how to create these new systems with this model in mind.

This is how we’re going to deliver the promise of AI to the corporate world. This is how we’re going to be able to deeply integrate AI into every aspect of our automation. We’re not going to be able to do it with ChatGPT 7. 0 or 8. 0 or 9. 0. We need this way or merging natural language and structured data embedded in inside the workflow.

We’ve already made those inventions. We know how to create that stuff since years. So that’s the model that we are basically proposing. And we’re calling that object messaging and intelligent objects. And based on this model, we’ve created one product. Which is really near and dear to our heart as software engineers, we built the kind of product that we always dreamed of by cobbling up JIRA and all these other tools that I mentioned, we made something that is a dream for ourselves, we had spread across the world and we just finished beta’ing this product a few months ago with two, three different sites and it’s the project management collaboration documentation, communication powered by AI and it needs to be all these because objects need all this information. They’re the ones who manage this information. So, I don’t have to file something in a particular folder. I just act as a human. I just act as a salesperson. I just log in and say, Hey, I have a customer who wants this feature and that object knows where to flow.

It’s a feature. It knows that it’s a feature. It knows It should go in front of this project manager and the project manager, if he decides to lower the priority, this object or my feature, that object immediately bounces back to me and says, my priority was lowered or I was put in backlog for two months.

So, the object knows and is aware who the stakeholders are and how to interact with them and how to get them involved at the right time and right place.

[00:22:41] Brian Thomas: Thank you for unpacking that and it just so happens I’ve been in healthcare for much of my career. So certainly resonates with me and a lot of our health care enthusiasts here in our audience, but love how you broke that down, how we’re going to really tackle this problem using intelligent objects. That is just awesome. And I appreciate some of the gems that you share with us today. Masoud, it was a pleasure having you on today, and I look forward to speaking with you real soon.

[00:23:05] Massoud Alibakhsh: Well, thank you very much for having me, Brian, it was a pleasure to share this wonderful technology with all my colleagues out there. So I would invite anyone who is interested and curious, we’re actually giving a free trial of our product that they can just go out to our website www DOT Omadeus, which is O M A D E U S .com, and just put your email in and we’ll give you a demo and you can try the product for free or sign up for the webinar.

We have webinars regularly and we talk about the subject actually have written a paper on this that is to be published by IEEE in November, and I will leave a link for you if you want to share it with your audience. I think that’s quite informative as well. Thanks for having me.

[00:23:44] Brian Thomas: Absolutely. And we’ll certainly put that information on our website when we publish this. So, thanks again. Bye for now.

Lauren Tickner Podcast Transcript. Listen to the audio on the guest’s podcast page.

Subscribe

* indicates required