Manuj Aggarwal is the Founder and Chief Innovation Officer of TetraNoodle Technologies, a premier data science, and AI consulting company focused on helping SMB & SaaS companies unlock the power of data and AI. We discuss ways to effortlessly get people’s buy-in, learning to automate AI processes, and how AI can help grow your business.
—
Listen to the podcast here
Bond by Sharing Your Stories with Manuj Aggarwal
Our guest is Manuj Aggarwal, the CEO of TetraNoodle Technologies, a data and AI consulting company that helps you harness the power of technologies like data science, AI, and blockchain for your business. Manuj, welcome to the show.
Thank you so much for having me. Excited to be here.
It’s great to have you. And your topic is so topical, so to say. It’s really hot right now. AI, I’m going to talk about ChatGPT and other AI applications. So we’re going to get into that. But let’s first start with your journey and how you went from a $2 a day job in India to running a successful technology business in Vancouver.
Well, it has been an interesting journey. I grew up in a small town in India and in those days, like this is 30 years ago, India is still, you can say it’s a developing country, but back then there were limited opportunities in smaller towns. So I, my career started at 16 working in a factory, as you said, and I was working six days a week and 12 hours a day. And I was making $2 a day and had some inspiration to do something different with my life. Specifically, I was going through some business magazines one day during lunch hour. And I read these stories of these tycoons who had built empires worth millions and billions. And I thought to myself, if these people can do it, maybe I can change my life as well. But at that time I didn’t realize what I was going to do or how it’s going to happen. But anyway, I found my love for computers and programming and I knew that was what I was going to do for the rest of my life.
And then after working in computers for a short time, I came over to North America in 1998 and I got a job quickly because that was the dot-com boom time and then I lost my job because of dot-com bust and then I lost another job because of September 11th then another one Gulf War so all of these things started to happen and then I don’t think there is a something called job security in even in North America but I’ll just do things on my own so I started my consulting company back in 2000 around 2001, 2002. And I started consulting with a lot of startups, consulted with the Microsoft, IBM, Pearson education. So I got to see like how small businesses work, how large businesses work, different industries, healthcare, education, logistics, and I got heavily into blockchain, artificial intelligence. So yeah, that was a journey. And today I have four patents in AI and I was just awarded, I recognize as a global thought leader in AI alongside with some really amazing people, world-class people. So that was my journey in short.
That is pretty much a Rex Rich story, the way we like it in America. It’s amazing. On this podcast, you always talk about some kind of business framework. And when framework and when we had our pre-call I was really struck by your framework, which has nothing to do with technology.
Yeah.
It was exactly the opposite. So, what is your management blueprint?
See, so obviously technology is something very near and dear to my heart but also I study a lot about human psychology and neuroscience and anthropology and history and all that. So once you start to learn about these things, you start to know what makes people do what they do. And one of the things that I had to learn the hard way in my professional journey was how to communicate with people, how to make them buy into my ideas, how to lead them basically. And some people who are listening may resonate with this that when you study in a professional like engineering, communication doesn’t come easy to you. You are just fact-based and there’s no gray area like black and white, right? So once I started understanding human psychology, I figured out facts are very bad. People pay least attention to facts. Even if they pay attention, they forget about it. They pay more attention to like how I make them make their mind sort of behave in that way?
And then I realized that there are two aspects to building these relationships, building, getting people buy in. First is ask good questions, because a lot of people live in their own head, and they are always focused on what is good for me, what do I need to accomplish. And one thing as I realized what great leaders do, they don’t really focus on what they want, of course that’s their long-term vision, but on a tactical level, they always focus on what do other people around them want, so that they can help them gain what they want. And the idea is that human evolution, human nature is such that if we help other people, it just comes back, whether it’s from that person or from a third party or whatever. So that is one principle to ask better questions to get to know the other person and their objective, their motivation. And when I say motivation, it’s not just, oh, you know, how much money you want to make? Do you want to get a promotion next year? But what is the end goal? Like obviously everybody wants to progress.
Great leaders don't really focus on what they want, of course that's their long-term vision, but on a tactical level, they always focus on what do other people around them want, so that they can help them gain what they want. Share on XEverybody wants to make money. But there is a hidden meaning behind making that money. Like some people want to buy a house for their parents. Some people want to go on a vacation with their spouse. Some people want to get their kids through college. So once you ask these deeper questions, now you get to know a person at a deeper level. And then what you can start to do as a leader is to actively start to help them facilitate that objective for them. An example could be if they want to go on vacation, it may not be that they need a raise. The next time you consider giving them a raise, buy them an air ticket and say, “Hey, I was thinking about you and here you go, I booked a vacation for you.” That will be much more meaningful to them than giving them a raise because they get what they want. Plus they know that you are thinking about their wellbeing, right? That’s one thing, asking better questions.
The next thing I realized was even if I share something meaningful our mind is not capable of remembering these facts and the thing about business and leadership and building something big is you know that you need to be remembered at the right time at the right place when they have a need of something that you’re offering or whatever if they don’t remember you the opportunity is gone. So how do you make it memorable? Again, if you learn about neuroscience, our mind has evolved to remember stories. So stories pack a lot of emotions, they help us remember our own past experiences. So basically, I started getting better at telling my story and not only that, but asking other people to tell me their story. And so in hiring, even in the hiring steps, when we hire people in our company, we don’t even look at the resume in the first step. We just say, “Here is my story. Listen to my story of going from $2 a day.” And then I go in a little bit of detail and depth of what challenges I faced. And then I tell them, “Now you tell me your story starting from where you started.”
And the magical thing is if there is a good fit between us, they automatically start oozing out their emotional challenges that they went through because something in my story resonated with them. It reminded them of something but if there is no good match then they just say “Okay I listened to the story what’s the next step?” So I know that not going to work no matter what. So that is the first step that makes the hiring process very efficient. First of all, because you can skip scanning like 90% of the resumes. The second thing is when they actually join the team, we already know each other as a human being. So the way of working, the loyalty, all of that is pretty much set from day one. Whereas in previous experiences and traditional hiring practices, what I realize is you hire a person and three months later you will find out whether this is a good fit or not. So this is a simple framework. And then there are extensions of this framework based on human mind and psychology and neuroscience that can be applied in other areas of business and I can go into those specifically. But once you start thinking from that angle, like a lot of things start to get, fall into place in a very unique way.
I love that framework. I love this idea of connecting with people, attracting the like-minded people who are going to be the right employees for you, going to be the right kind of clients. I find that I’m more resonant people who think like me, who have a similar story. Maybe they run a similar business to what I ran at the time. I typically professional service tech enabled firms. I love those because they resonated with my story. I resonated with their stories. I get excited about their business. So they see it, it drops off on them. And definitely I co-create with this idea and I like the intentionality of it, sharing your story. I share my story. Let’s see if we resonate, if we resonate, let’s do business together. If not, that’s fine. That’s okay. Let’s find someone else who is more emotionally bought into each other’s ideas. Love it. So I guess you could spend the whole podcast on discussing that and fascinating topic. I think it would be interesting. However, there’s so much more I want to ask you about and especially now with ChatGPT blowing up and you’re an expert, you’re recognized as a thought leader in this field. So I have a few questions for you. So let’s start at the beginning. What does AI mean for a medium-sized, small to medium-sized business? How do, how should we even think about AI? It’s such a big idea, but it’s very slippery. how do we grab of it.
I love this idea of connecting with people, attracting the like-minded people who are going to be the right employees for you, going to be the right kind of clients. Share on XSure, a good great question so at a fundamental level AI is basically a pattern recognition machine pattern recognition in a sense that it can recognize very minute and very complex patterns. So our mind is not able to recognize like very complex patterns and I’ll give you an example. So let’s say if you touch a hot stove, you will immediately realize this is not a good thing to do. And so that is a data point that your mind remembers and now you have recognized the pattern that touching the stove is not a good idea. Let’s say you are in Antarctica, you’re wearing heat resistant gloves, you’re outside, it’s negative 50 degrees centigrade and wind is blowing and now you touch the stove, it actually feels nice, it’s warm. And now your mind recognizes there are variations in this pattern, right? It’s not just don’t touch the stove, but what are the conditions? What are the environments? So there are like five or six parameters I mentioned there.
And now our mind can recognize that but now extrapolated to like bigger problems climate change or predicting the weather and there are millions and millions of parameters there and our mind is not capable of recognizing these patterns. But if you feed all this data into a machine and ask, “Okay, tell me what’s going to be the weather in two weeks.” It will be able to see those patterns and tell us what’s going on. So that’s the fundamental thing. Now bringing into the business sense, first of all, we need to start looking at it as a very intelligent employee. So let’s say you hire an executive who has a degree in MBA degree in law degree in medical science degree in multiple other degrees how effective and intelligent and you know, what kind of outcome you will be able to get from that employee that is AI today, right and then on top of that this employee also knows how to do things automatically at the speed of light without error. I can that is what AI is right. So let’s say a small to medium-sized business and I hope that concept is clear like what AI is. Yes.
Yeah, I love it. Okay. Love this idea of error free and automatically at the speed of light. That’s super exciting.
So now how do you apply this? How what does it mean for a small to medium-sized business. So what I tell people is that if “I cannot sit here and tell you apply it here, apply it there.” Obviously generative AI and chat GPT is you can create content, tell stories, ask better questions, but also think about what is the most biggest challenge that you are facing in your business. Every business has challenges. Some people have sales challenges. Some people have marketing challenges, operational challenges, whatever it is and then look at the most important API that you’re tracking in regards to that challenge. Meaning if you’re tracking revenue, if you’re tracking productivity of your employees, if you’re tracking growth in territory size of whatever that is and then you can say, okay, I need to maximize this API. So let’s say you you’re tracking revenue, which is generally what businesses are all about and I think the biggest challenge in increasing my revenue is that my team spends too much time prospecting for new clients.
That is the definition of the problem and now we work backwards from there and say, “Okay, how can I use AI to intelligently find the right prospect? How can I automate a process where the prospects sort of discover me and come into my door. How can I personalize the approach so that the prospects know that we understand them very well.” So that’s one example, right? Another example could be, oh, we are running a manufacturing unit and the quality of our stuff that we manufacture is degrading. People are not happy. So how do I make sure that the quality is maintained? That’s a separate problem. But again, AI can come into the picture where we can say, “Okay, you know, we can install some cameras on your assembly line, which can take very, very rapid photographs. And those photographs can be analyzed by AI, tell you, this piece is defective or whatever it is.” So every problem has a different solution. But you can see how, if you have a super intelligent human being with all these capabilities, you can actually get it to do whatever you want it to do and solve your problems with low hanging fruit, the biggest bang for the buck and all that.
Every problem has a different solution. Share on XOkay. So I get that. So I can think of AI as a super intelligent employee and I can ask questions. So ChatGPT is all about that. It’s asking questions and the better we can frame the question, the more accurate answer, the deeper answer we’re going to get. And it’s basically for prompt engineering, how to ask smart questions, the better we can get to that. I read that prompt engineers who can ask smart questions can make $200,000 right now. But that, but still there has to be someone there who asks the questions. The other aspect is also very intelligent, interesting, which is what you said that how do you automate things and have that digital person, this robot, actually solve problems error-free at the speed of light? So if I’m a small business owner, I say, okay, I can ask questions. So that is a great start. I get a lot of good answers and hopefully I am somewhat knowledgeable in the field so that I’m not going to allow it to hallucinate me, BS answers. Okay. So I’m there. there, but then how do I turn it into how do you harness it into actually doing things without me? How does that work? How can people visualize that process?
Yeah, so this is where you if you’re doing something more complex, then you may want to work with an expert like myself or somewhere else where they can break down that process into smaller steps. Every business is a series of steps of series of workflows. You get an order, you pass it on to the onboarding team, it goes to the operations team, you fulfill the service, blah, blah, blah, right? So the idea is not to look at this behemoth as AI is, okay, what the hell do I do with it? But it’s like, okay, what are the steps I am doing right now? If this step involves human being, can I replace that? What is that human being doing? Is it a repetitive thing they’re doing? What kind of decisions are they making? Is it like very complex decision or is it like just checking? Okay, is it inputting some data or checking whether there’s a crack in this?
Attendant permission.
Yeah. So if we can find those simple patterns that can be trained for everyday eye very quickly and then you can automate that process. The next one could be human to human interaction, which again, chatbots and all that, they are getting better, but you may not be comfortable replacing that just yet because that is the face of your company, right? So you can say, okay, leave this step as human. We don’t want to touch this. Now let’s go over to the next one, fulfillment and warehousing or whatever. In warehousing, in even medium size, small size businesses, inventory control is a big problem because it costs a lot of money. So you can start to use AI or data analytics to see, okay, which products are moving faster, which products have seasonality, which products may not be as popular, maybe like certain, let’s say you are a clothing manufacturing company and you manufacture t-shirts. So which color is not in fashion these days. You can recognize all these by analyzing data and then you can bring it back into your business, right?
Okay, Manuj, I get that. My question is more, and I think it’s very important what you said. My question is, the next question is more practical. Help me understand how this visually looks like. So does that process have to live on the digital sphere for it to be automated or physical steps can be automated. So where people are doing things physically or going here or there or pressing buttons. So how can this be, does it have to be digitized as a process, have to be digitized first and then AI can be applied or there is maybe AI can figure out a way and then there are other tools that can be put there.
You know, no, you don’t have to digitize everything. So I’ll give you another example of it, right? So in a big car manufacturing companies, now I’m using a big company example, just to demonstrate it. You’ll see, you may have seen videos of robots, like there are lines on the floor, they are guiding lines for the robots to move from one place to another. In the past, that used to be a human activity where they had to go from one place to another. And now that same thing can be automated, utilizing multiple disciplines of AI. So AI, there is a discipline called computer vision, which helps the computer to see just like humans can see. There is another one where it can listen to sounds and things like that. There’s another one. So there are multiple of these disciplines which you can stick together and say now this process or this machine or what have you will behave similar to a human being.
So it’s computer vision. What about natural language processing connect the computer vision to AI?
Yeah, natural language. So let’s say, and again, I think natural language processing as a technology will be outdated given how much ChatGPT has made progress. So natural language processing was, you can call it like a baby in front of what ChatGPT has done, right? But natural language processing could be where, take an example of going to a drive-in window at McDonald’s, right? So we generally say, “Give me Big Mac.”And they’ll say, “Oh, will you like to supersize that? And will you like fries with it?” Whatever it is. So now you can imagine that those systems are available today, by the way, which can recognize your speech in real time and then intelligently ask a question. And as long as it’s not like a deep deep conversation you can have a chatbot sort of ask relevant questions and then I recognize that this is what you want to order right now that order can be placed into a machine which starts to process this order and actually start to deliver that order in the next window completely automated completely without human intervention. Those systems are available today.
So natural language processing. I thought it was more the processing of the words that was recorded by this computer or heard by this computer like the same way as the computer sees computer here’s and then the data goes in and that’s natural language processing turning it into intelligible information that I’m going to pass it by AI because it’s our trucks.
So natural language processing is exactly what you described like it listens to the or listens or processes text and then tries to understand the meaning behind it. Like is the customer angry is the customer annoyed is the customer and so natural language processing was like if an elk then if and else logic if you find this word with the technology that that GPT is using, it’s a large language model. So it can actually sense the sort of the context not because of logic, but because of how it knows about how we as humans communicate. So to describe you a little bit in more detail without getting into more technical jargon. So natural language processing is generally based on pattern recognition where we apply logic and we look for certain words large language models what they have done. They have taken entire data set of almost the entire written text of humanity. And what they have done is they have looked at each word and now they have recognized what word comes after which word. So it in human language when we teach our kids we say A for Apple B for boys that the level of depth large language model has gone into it knows how we communicate with each other and then it can reverse engineer in real time as we speak to each other that okay I and I recognize this pattern this pattern happens when people are annoyed or when they are happy there is no logic needed to train that model. Does that make sense now?
Natural language processing is exactly what you described like it listens to the or listens or processes text and then tries to understand the meaning behind it. Share on XYes, I understand. Okay, so there is computer vision, there is computer audition, I guess. Yeah. Natural language processing, then you have the large language models, which basically have all the humanities written text and they have requested patterns of what comes after which, what is the context. I get it. Now, GTP uses this. It’s a pre-trained generative model, a GPT, generative pre-trained transformer, right? And so that’s GPT. Now how can a business, let’s say a business comes to you and says, okay, Manuj, help us use AI, help us make our business more automated, more efficient, more cost efficient. What do you do? How does the engagement look like? What can you how can you help?
Yeah, we go through an exact same process as I described earlier, we sit down and say, okay, let’s understand your business. Tell us what is the outcome you’re looking for? What is the biggest challenge you’re facing in that so we can take a hypothetical scenario and we can say, okay, we want to maximize our revenue. And we are getting really bombarded because our sales are stagnant and competition is very stiff. Let’s say now what we need to do is we need to start collecting data. What is the competition doing? Why are sales flat? Is there is the market shrinking or the market is growing but this company is stagnant because if the market is stagnant pretty much we are closed unless we provide a new product or a new service. So that’s data analytics. That’s this market analysis data analytics. Now if we let’s say we determine okay, the market is growing market is healthy.
There’s a demand for the product and now more competitors are entering into the market. Then we say, okay, how can we differentiate your product or your company or brand online or in the places where your audience is found. So this is again, nothing out of the ordinary again market research, but once we start to get to the human aspect of it, but let’s say we find out. Okay, these people are hanging out in a discord discussion channel. And now I know that okay based on the data that we have collected, you know, this is a younger audience because this code usually is popular with younger audience. And now what are the trigger points for those younger audiences that we can start to harness? So it could be something to do with the pop culture. It could be something to do with fashion.
Even though this is a B2B product, we can start collect data on their interests. And now we need to start creating some sort of a link between their passion and what this product is doing and that can be that we can start doing with AI that we can start to rapidly scale and get more visibility by capturing the attention of these people by using AI because doing it with human intellect it’s very very difficult. So I hope these steps will be going from their target to breaking it down into steps, collecting data on each one of them. And not everything is AI because some of it is just regular market research, some data analytics and stuff like that. But wherever you think that AI can help, we put AI in there, computer vision in there, in manufacturing and security, computer vision is very, very prevalent. In some other areas, computer vision is not required. But this is how we work. Does that help?
Yeah, I think it does. So I understand the data analytics piece. I understand you, you know, you dig in to understand what is the business challenge and then collect the data to analyze it. And then you try to get into the emotional needs of the target market and you use AI, connect the dataset with this emotional needs is how I interpret it and essentially iterate to find the best combination that perhaps for the human mind would be really difficult to synthesize down.
Exactly.
Too much information. And then you can basically serve the kind of ads to the audience that will create the response and the product.
Yeah.
That’s fascinating. Yes. Anything else?
I’ll give you another example if you want in a different domain. So we actually executed a project for a higher education industry like universities and colleges. So there’s a big problem in these universities where when the students enroll into these programs within the first two years, they drop out because they find that the degree program they enrolled in is of no interest for them or it is too difficult for them. So this results in a huge loss for these universities like up to like 500 billion dollars and society loses out on a chance to get a graduate, right? So what we did was with AI, we collected data from hundreds of thousands of students and we figured out based on their SAT scores, their aptitude scores, their interest, what courses are of interest to them and what are they able to finish. So we created that model and then when new students enroll, we get them through this model and we say, okay, based on your past history, you should be taking these courses. And now these courses are a better fit for them. And the result was if every student that went through that, they took 20% more courses than average and they, their dropouts really declined. Right. So now you can see like how AI can really tap into the collective intelligence of human beings and then start to help other people as well.
Yeah, that’s fantastic. Well, I hope that the next election is not going to be what AI is. I’m just scared that the quantifiers are going to tap into our emotions even deeper and have even better way of manipulating us. Hopefully it will not happen.
Well, yeah, I’ll say one thing like it has been happening so far, but I think the good thing is now the power is in the hands of everyone. So I think it will happen less.
Okay, well, that’s probably a topic of a podcast series. But I hope your optimism is warranted. Our listeners would like to learn more about tetra Noodle Technologies that are your business about maybe they want to learn more about what you do, what connects with you, where should they go?
Yeah, you can check out our website, tetranoodle.com or my website, personal website, manujaggarwal.com. Or the easiest thing could be just connect with me on LinkedIn. I’m very active on LinkedIn.
Yes, you are. All right. So definitely check out Manuj Aggarwal, the CEO of TetraNoodle Technologies on the website LinkedIn. Manuj, thank you so much to come on the show and to share your wisdom and knowledge of this evolving AI, HGPT area, which a lot of people are very excited about it and engaged with it. So thank you for coming. And those of you listening, if you liked what you heard, then stay tuned because every week I’m bringing an entrepreneur who has a substantive business and have a mass information and frameworks that you can benefit from. So stay tuned. Thank you so much. Thanks.
Thanks very much.
Important Links:
- Pinnacle: Five Principles that Take Your Business to the Top of the Mountain
- Stevepreda.com
- Manuj’s LinkedIn
- Manujaggarwal.com
- Tetranoodle.com