Anmol Satija -Hi, everyone. So before getting started, I have a small scenario for you to imagine. Think about an AI system that doesn’t just follow your orders but sets its goals, learns on the fly, and takes action all without minimal human intervention. Sounds like an ideal situation, right? This is exactly what Agentec AI promises.

So hello and welcome back to the Unthinkable Tech podcast, the go-to source for the pulse on technology shaping our future. I’m your host Anmol Satija. And today I’m joined by a very special guest, Jesse Angelen, who is the founder and CEO of Rapid Innovation. Jesse is a visionary in the AI space and holds deep technical expertise and a relentless passion for driving innovation. So today we’ll be covering everything exciting about agentic AI.

How it is different from traditional AI systems, and why it is poised to be redefining industries. So buckle up as this is one of the conversations that you don’t want to miss. So hi, Jesse, welcome to the show. How are you?

Jesse Anglen: I am doing fantastic. I live one of the best lives ever because I get to do what I love to do every day, which is not what everyone can say. so most days are really good for me.

What is Agentic AI? Understanding the next evolution in artificial intelligence

Anmol Satija: Sounds great. So let’s just get into our conversation. So to kick off things, let’s start with the very basics. So can you define what Agentic AI is and explain how it differs from traditional AI systems and GenAI? What exactly makes it so unique?

Jesse Anglen: Yes, so I will say agentic AI as a term is a moving target because people are still kind of defining what an agentic system really is. And so I’ll kind of give you the high level and then break into some of the specifics at least as to how I see it and how I think a lot of industry leaders are looking at it.

Really, so the first thing you need to think of is an agentic AI as a synthetic brain or maybe AI as a synthetic brain, right? So we’ve got these large language models and in a sense, they’re able to mimic human thinking, which basically just means that they’re capable of some form of autonomous thought. They’re able to mimic planning and they’re able to pretend or mimic taking action. so agentic AI takes advantage of all of those things basically by creating a system of agents that is, instead of being very task-specific, right? So if you think about the AI landscape, there’s AI out there that can write words. There’s AI out there that can generate pictures, generate video that can speak, that can do all these different things, right? And they’re very task-specific or niched in what they’re able to do, which means that a human being is required to kind of chain all of the different aspects of what an AI can do together. Generative AI basically just focuses on creating an output for the most part. And so,

What’s unique about agentic AI is that it’s able to set and achieve its own goals and then work together with other AI agents or models to accomplish those goals. As it’s going through that process of accomplishing that goal, it can actually learn things, maybe learn something from its environment, learn something based on those token inputs, and make decisions about what it learns, store what it learns and then basically act on those things. And so you might string together an agentic system that, gives you a really basic example, right? So you’ve got an agent that does research. So maybe it uses perplexity or some other model. Maybe it’s just literally searching the internet for stuff. It gathers that information. It synthesizes it in some way. Then it passes that information to a copywriting agent.

The copywriting agent takes that research and turns it into a blog article. It passes that to say a manager AI that understands the business objectives that you have. It takes and provides feedback back to the copywriter agent. And then they all work together making decisions and communicating with each other to accomplish a specific goal, which in that case would be writing a blog based on research that it finds on the internet. So that’s a simple idea of what an agentic system is.

Anmol Satija: Interesting. I think that’s fascinating and it’s a kind of a significant leap forward. Agentic AI seems to be more like a self-sufficient thinker, I believe. And as you said, it doesn’t just react, but proactively decides and adapts.

Anmol: So, you know, this opens up a whole new realm of possibilities. So let’s dive a bit deeper into this topic and see how it works in real-world scenarios.

So as you just mentioned also, we often hear about AI agents as the driving force behind agenting AI. So could you shed some light on their role and how they enable intelligent automation in practical scenarios?

AI Agents: The building blocks behind intelligent automation

Jesse Anglen: Yeah, basically think of an AI agent as either an employee or the operational unit of the agentic AI architecture, the agentic AI system. And so it’s either like it’s the input or it’s the input and output of a system, just like a person would be.

They’re going to look around at their environment. They’re going to understand they have a task. They’re going to analyze their options. And then they’re going to act without any direct human input. And so in theory, you can give it any sort of a task. It’s going to create, or you can give it any sort of a goal or a task. It’ll create all the tasks associated with accomplishing that. In some cases, it might even hire, so to speak, or fire other agents by spinning them up, right?

So the system will create another agent if it’s missing someone on the team and then learn and improve as it’s going. And so in real-world like situations where you start to see these things take place would be kind of the obvious one, content creation. So people creating videos, people creating blog articles, web pages, people writing code, those are often the newer technology that’s coming out. Those are often agentic systems, but you can also see people starting to explore using them in like in other areas. like dynamic routing of vehicles, right?

So you’ve got an agent that’s looking at the traffic patterns where, you know, pulling that data from Google, trying to figure out what is the actual fastest way to get from point A to point B and updating that in real-time. So it’s like having a co-pilot in your car with you navigating as you’re running around, trying to find the best route from point A to point B.

Next, you see it being used where in customer service. So areas where you’re having a conversation with an AI agent, but in that conversation, there might be five AI agents. One of them has access to your billing information. One of them has access to all of your past customer inquiries. One of them has access to all of the FAQs from the company so that it can best help you. And they’re basically feeding that data back into the conversation so that as you’re talking to the AI, you can have the best experience possible.

And then you’re seeing people innovate in areas like monitoring equipment health, right? So you have an AI agent watching the status of a machine on a manufacturing floor. And they’re basically trying to figure out like, what’s when’s the best time to do maintenance on this machine that will reduce the most downtime. can actually, because it’s an agentic system, it could look at, let’s say you’ve got a scheduled maintenance on Wednesday.

But if you shut the machine down on Wednesday to do the scheduled maintenance, you’re not going to hit a customer order. And so then it can look at when the last maintenance was and they can start actually making intelligent decisions about how to do day-to-day operations in your business in a way that human beings just can’t.

Anmol Satija – Right. It’s really incredible to think that these agents are working as problem solvers behind the scenes, optimizing processes around the clock. You know, this shows how diverse the applications of Agentic AI can be.

Anmol: So speaking of applications, now for organizations exploring where to begin with Agentic AI, what are some of those practical and impactful applications to say or how do these align with the priorities of enterprises today?

Jesse Anglen – So it’s interesting because I have conversations with people about agentic systems all the time. And because we’re in this space where we’re just now exploring it, I would say that there’s a lot of creativity going into where those practical starting places are. One of the things I think that will have the most impact as people start implementing is going to be customer service and dealing with customers, right? Because there’s only a certain amount that a human brain can actually hold. And it takes a significant amount of time to train someone to actually be really proficient in the customer service area. Well, if you’re dealing with an agentic AI, it can literally hold all of the information. You you could have a fine-tuned model that holds all of it. You can architect these so that it knows everything it needs to know to do its job and then have a very human-like conversation with somebody.

And I would say in some cases with some of what we’ve built, even to the point where 95 % of the people that are talking to this agentic system, they have no idea they’re even interacting with an AI. They think they’re talking to a human because its response times aren’t very latent, it’s creative in how it responds, but it has access to all of your company’s data, can help with any problem from any department, right? So that’s gonna like fundamentally change things as people start doing stuff. So you’ve got customer interaction, then I would say the creation of marketing material is going to be another kind of offshoot of that, where you’ve got, like, if you think about it, I mean, you’ve been in marketing, I’ve been in marketing, right? You have these big teams of people that are trying to create all of this content. Well, if you are a business owner, and this is a little brutal to say about the people that are working, but if you’re a business owner, it is much easier to deal with an agentic system than it is to deal with people, right?

Like the agentic system doesn’t call you up and say, hey, I just got in a fight with my, you know, with my wife, or I just had, my kids just got sick or I’m tired or I didn’t have my coffee this morning or whatever. Like they just do their job.

Enterprise applications of Agentic AI

And so when you build systems to create content or do different things like that that adds value to your clients or to your customers, that’s going to be a big use case. But then I think on the more practical side of things, you’re going to see it in supply chain automation because it’s just going to be more efficient than people. You’re going to see it used in predictive maintenance because you’re going to see it be way more, it’ll make less mistakes than people.

So, there is this current realigning with the new world of agentic systems happening in enterprise right now, simply because it’s reducing operational costs and improving efficiency. When you don’t have to work with people who are expensive and inefficient, it just helps. Also, in areas where people are implementing it, the customer service is getting better, not worse. Because it’s not like you’re talking to the…

Verizon robot, right, that is absolutely unhelpful. Press one to do this. You’re talking with something that is almost indistinguishable from a human. so at the end of the day, what’s happening is that people are implementing these systems, and as it increases efficiency, they’re just making more money. And so…

That’s going to be the driving force behind doing this, right? If you can implement an agentic system that costs you, say, $300,000 to build, and it removes 10,000 customer service workers, all of a sudden, the ROI is just out of this world. And so I think that’s where we’re going to see a lot of the driving force behind what’s happening.

Anmol Satija – Right, right. So those were some really exciting examples, I would say. And you know, your insights made me think that Agentic AI isn’t just another tech trend. It’s also about solving real problems and driving measurable outcomes out of it, right?

How to implement Agentic AI: Strategy, pitfalls & practical frameworks

Anmol: So, but as with any powerful tool or new trend that comes up, implementation isn’t one size fits all approach, right?

So let’s talk about how companies should approach this strategically. So as per you, what approach should companies adopt for the implementation and what critical consideration should they address upfront?

Jesse Anglen – Okay, so I’m gonna go like to answer the question, I’m gonna go back to like business 101, right? Like when you are operating a business and you’re thinking about change, like the first thing you wanna look at is what problems do you have? And then what technology exists to solve those problems? And so you have to be really specific, you have to define the problem and then you have to figure out whether or not AI can produce a result that you can measure.

And so for instance, like just an example from my company, when somebody would call us up and they say, Hey, I have this piece of software, custom software that I want to build. We had a process where we would then go and we need to discover what that person wants to do. Right.

It requires a large team because they need a business analyst. need a senior backend developer, a senior front end developer. I need someone from my DevOps. I need a specialist from the blockchain department or a specialist from the AI department. And they all have to work together with the client to understand specifically what he’s trying to do, what his needs are, so that we can figure out how long it’s going to take, what architecture, what tech stack, what team we should put on it, and all this stuff.

Well, that process takes quite a while. In the old world, before agentic systems, I would say that it took between three to five weeks to do that whole process with a team of five people. Working, let’s say, quarter time, right?

So a significant amount of time, but not full time, like quarter time, because there’s a lot of back and forth. Well, when we created our first small pilot in that area and said, okay, can agentic systems offload a lot of this work from my critical people so they can work on projects and just let the agentic AI did it. And we followed through with that testing, what we discovered is that we can do that three weeks worth of work in about two hours.

And so then the only thing we’re waiting on is for the client to get back to us. And so we can get through the entire discovery process, even waiting for the client to do things in about a week. But it only takes, I mean, it takes zero time from my guys to do that. And so you identify a problem, identify an area where you can create a solution and then start piloting it, you know, build something small, see if it works and then start stringing those systems together.

And so I would say another couple of things that you ought to pay attention to is for all agentic systems and really AI in general, there’s this garbage in garbage out consideration, right? So if you give it garbage, it gives you garbage. And so you want to make sure that you have a way to collect good data to feed your agentic system so that it can produce a good result for you.

The other thing that I think larger enterprise are going to have to deal with is the ethical side of this and the morality of it. I’ve got a client right now that we built a system that could, I mean, in one month with one developer, we built a system that could replace 8,000 employees. Well, as a business owner, do you want to do that, right? Like, is that an ethical decision? Can you do that and actually, you know, maintain your company boundaries and all that kind of stuff. So establish those guardrails before you get into it. So you kind of know what you’re going to do, what you’re not going to do. And then also like what decisions should AI make? Like should it be making decisions where it can discriminate against different people? Should like, so be thinking through where you need like humanity, a soul, a heart, like someone who cares to be involved in that decision-making process.

And then just align it with your business goals. Make sure that the AI actually understands what it is you’re trying to do when you build these systems. It isn’t a one size fits all generic thing where you can just do whatever you want. Oftentimes you need to make sure it’s aligned with what you’re doing and where you’re going. And then right now, one of the problems that I see is that the tech division of companies is very focused on agentic systems, but oftentimes it’s very fragmented from the business side of stuff. So pulling those two departments together where you’ve got the operators and the business portion of what you’re doing and then the technical side of what you’re doing, make sure that they’re actually aligned because it’s never been more important than when you’re actually building this kind of logic into your business.

Anmol Satija – Yeah, definitely. Definitely. I think that’s an excellent advice that you gave, starting small, thinking big, and staying up all aligned. So it’s clear that thoughtful planning is the key to success. But now let’s turn to one of the most exciting aspects of Agentic AI. I’m talking about its potential to augment decision-making. So how exactly can businesses leverage that?

Augmenting decision-making with Agentic AI in the enterprise

Jesse Anglen – Yeah, so one of the like, one of the projects that I’ve wanted to do for a long time that eventually we will get to and finish is I’d like to create a synthetic version of my brain that makes decisions as the CEO of Rapid Innovation, have that AI fire me in a board meeting and then start making decisions. Now, I don’t think it’s quite there yet. I don’t think it, I don’t think we’re actually going to, we’re not in a place where we can make that happen to that extent, although we are getting there.

But here’s the biggest advantage of that, right? Is that if you know, I’ve got several hundred employees that work for me. Let’s say that one day all of my employees have a question that only I can answer. Well, how long does it take me to go and do that? Right now I’ve got a lot of options. I can set up a town hall meeting and I can do all this different stuff. But it is so inefficient, the bottlenecks for decision making in business is so inefficient.

And we don’t even know it because we’ve been living with it forever, right? It’s very difficult to get information from the sales team about pricing over to the operations team about the cost of living that’s going up. And the HR team is they’re hiring people because number one, it’s hard to get that information to one person’s head. And then two, it’s hard to even understand what kinds of decisions you should make to do this. Companies hire huge teams of analysts to look at all this information and then start actually giving suggestions or making decisions with that.

Well, from an operational perspective, AI can just do that. Like it sees all the information, it creates the inference, it can just give it to you. All of your employees could go to an AI that understands your company. They could all ask a question at the same time. They all get an answer at the same time. They can all ask follow-up questions. And so you are decentralizing
the decision-making intelligence by creating that synthetic intelligence that can start to do that, which then gives you this opportunity to really have strategic data-driven insights inside your company. You can start making better decisions faster. I think everyone knows if you can make better decisions faster, then you’re going to outperform your competition. You’re going to be more successful. It improves everything.

And so like practical examples of this might be like retail in retail, know, like understanding that there’s a pricing spike or maybe there’s a there’s a huge fall off of demand. Well, then what do you do and how do you do it before your customers? And so, I think I’ll stop there. Like, that’s a that’s kind of a high overview answer to that question.

Anmol Satija – No, you explained it pretty well. So those were really interesting insights to know. So now let’s come up to our last but not the least part. So looking ahead, how do you see the landscape of Agnetic AI evolving? So what advancements and challenges are on the horizon according to you?

Jesse Anglen– So Ilya Suskovor, basically is one of the smartest people in the world, he said something very interesting the other day. think Sam Altman echoed him. Or maybe Sam Altman said it first. I can’t remember. But he basically said the time is coming where we’re going to have agentic AI and maybe combine possibly with AGI or artificial general intelligence.

The future of Agentic AI

And when that happens, we’re going to start seeing companies that are formed that are unicorns with one employee. So one person will have an idea. know, Anmol is going to go out into the world and she’s going to go, you know what would really help the world is doing this thing. And then you’re going to go have a conversation with an AI. The AI is going to write all the code. It’s going to create all the marketing. It’s going to do all of the work. It’s going to create the supply line. It’s going to organize the manufacturers, it’s going to do all of it based on your human creativity. And over the course of, you know, what I think a unicorn is in under five years, you have a company worth a billion dollars. And so over the course of that five years of you, just you working with the AI, the AI is going to build you a billion-dollar company. Like that world is coming, right?

And so in order to get from where we are today, because right now, honestly, if you look at agentic systems, they’re not that smart, right? They’re not going to help you accomplish some crazy goal like that. And so we’re going to start to see integration, right? We’re going to start to see this architecture where everything begins to work together, because it’s kind of just a big experiment. know, one of the things was really interesting when generative AI first came out, and I was trying to tell people about it because I thought it was so amazing.

Everyone’s like, that thing that writes poems. I’m like, you’re staring some of the most incredible technology that humanity has ever seen on the face of the planet. You’re staring it in the face. And the only thing you can think of doing is asking it to write you a poem? This makes no sense at all. But we’re doing the same thing with agentic systems where the focus isn’t as broad as it needs to be, and it’s not as connected as it needs to be. Because as you get those integrations in place, right, like it’s able to be very multimodal, understand different kinds of information, create the inference off of that information and actually do something with it. I mean, it gets smarter basically, right? It knows how to, like a human, take all the different pieces and do something with it.

It starts to actually learn from what it sees, creates its own data, and adds that to the vector database so it can pull from that data in the future. Basically, like self-learning algorithms. Like that’s where we’re going to see this just absolutely take off. And there’s so much research and so much work being done in this area that I think over the course of the next five years, it’ll fundamentally change the world. And it’s not like it’s without challenges. Like we have to figure out ethically what to do.

Because there’s a lot of people in the world that work. Like lots of people, billions of people go to a job and they make money from their job. Well, what happens when my client decides to implement that customer service bot that could be created from one developer over the course of a month and 8,000 people lose their job? Like there’s some ethical considerations. If AI is making decisions about who gets a loan and who doesn’t, like how do you deal with that?

And so I think there’s some of that that’s gonna have to happen. Then you’ve got the public, right? When I talk to people about AI and agentic systems, especially just like my friends or my neighbors, people that don’t know a ton about it, it’s very interesting. One of the first things almost everyone says to me is they say, well, that’s really scary, right? Because new technology just looks scary.

And so how do we get people to a place where they actually see these systems work? It’s not quite the scary black box that it is right now. And they start to build confidence. And to be quite honest, I don’t know how we do that, but it’s going to have to happen. Because at the end of the day, mean, really what it boils down to is this. Like when AI is working with you or an agentic system is working with you as a partner, and it starts to augment human ability, right? It makes you a better podcast host. It makes me a better CEO. Like it’s starting to, making my employees better at coding, better at HR, better at all of these things. It’s almost, I think, impossible for our minds to wrap, like wrap our minds around the concept of what that does to humanity, because it’s a fundamental shift in the way that planet Earth and human beings have ever functioned.

Like we’ve always relied on human labor, human intelligence, human decision making. Well, what happens in a world where we’re not relying on human labor, human intelligence, know, human problem solving? It solves a lot of problems. It creates a lot of problems. And so I’m really excited for what the future holds, I think it’s going to be amazing. I get to watch entrepreneurs do epic and amazing things every single day in the AI space, which is super, super fun for me.

And like part of one of the things I’m really passionate about is making sure that people know that this is coming because most people think that think that large language models write poetry. That’s what most people think. Or you can create a fancy song or make a nice pretty picture. People don’t realize that it is.

Especially with the correct architecture and the correct way of thinking, they don’t realize what it can do, that it is absolutely transformational in its ability to solve problems. so, like that’s where I think like we need to just go as an industry, people interested in agentic systems, we need to get the message out there as people that are building, we need to do a good job, right, of educating people, showing them how this magical stuff works.

Anmol Satija – Yeah, definitely. Yeah, I think the thinking is gonna change. even, mean, this episode could do that work, right? We have shared a lot of interesting information that can help them think beyond poems and writing songs. Right? So it was incredible to, you know, to get to know so much interesting facts about Agenetic AI from tackling global problems to revolutionizing industries. While challenges like you mentioned, ethics and trust are critical, it is clear that the potential impact of agendic AI is nothing short of transformative, I would say. So it was a great conversation with you, Jesse. And I just loved whatever you said. And thank you for showing up. Thank you for joining us for this episode.

Jesse Anglen – Not a problem at all. You’re a great host, and so I appreciate your time. It was fun.

Anmol Satija – Yeah, yeah, for me too. And thank you to our listeners. If you are curious about how Agentic AI can revolutionize your business, now is the time to explore. Thank you all for tuning in and until next time, keep pushing the boundaries of what’s possible. This is Anmol Satija, signing off.