Latest / The Unpack Show with Anandh / Ep-16: Harveer Singh | Co-Founder & CIO, Rizz Wireless
Transcript
- TheUnpackshow: Cool. Hey everyone. welcome to the Unpack Show with Anand, where we unlo unlock and unpack a lot of ideas about leadership and philosophies and innovation. I'm your host, Anand. you know, this topic is an interesting topic because it's close to my heart and everyone is dealing with this on a daily basis. we are living through one of the most significant technology shifts since the internet, right? So Which is nothing but AI, right? So anyway, if if you don't have an AI in a conversation, it's not a conversation, in my view, right? So the topic for today is empowering people in an AI-driven organization, right? That's the topic for today. you know, if you look at technology, technology has never been the hardest part of the change, right? Throughout history, organizations have successfully implemented new systems but struggle to l help people adopt them, right? That's the biggest thing. so if you look at AI, AI is no different. Success won't be determined slowly by the sophistication of an organization's AI capabilities, but it'll depend on whether leaders can build trust and also develop new skills and foster a culture, right? Where the continuous learning and empowering employ employees to work confidently matters, right? So today I have a special guest, and he has done a lot of this. work in his previous life and he's a co founder. I will let him introduce in a moment. without further ado, Harvey, why don't introduce yourself and we can dive into the topic. Harveer Singh: Absolutely. And thank you, Anand, for having me on your show. I have listened to your previous... Someone is at front door. I love when Alexa interrupts us. I'm not going to say cut, but let's... TheUnpackshow: That's that's AI. You need to have an AI, right? Harveer Singh: Look, it's life. And sorry, Diversing. I never let my kids, you know, never say no to them if they want to interrupt it. That's totally fine. I think we are living in an age where I TheUnpackshow: Yes. Harveer Singh: would miss them when they're not around. So it's okay, you know. hearing the voice of Alexa that somebody coming to your home, it's actually a pleasant blessing, so I'll take it. So about me, first of all, thank you for having me on your show. And I viewed, I've seen the past and we've discussed about this topic. I'm truly excited to share my findings. I'm actually doing a lot of work in this space and we'll talk through how some of the stuff that people are envisioning. I think if you don't get your hands dirty into it, it's hard to speak at the forums that you command. So I'll be going through what I have been testing and how I have been testing, et cetera. But as a background, I've spent about 25 odd years in data, in technology. I was a chief data officer on the consumer bank side of a tourist. Prior to that, I had led Western Union as their head of data and their chief data architect. I got to see a lot of the global data movement, which also inspired my book, which is the when data moves. And I'm blessed that people are giving a great reception. It talks about how we should think about technology in a world where everything essentially will become AI. If we don't have a human centric view, then the AI is just going to look like us. the empathy that we have, it's going to be translated to the empathy to AI as well. So, and we'll talk some of that in today's topic. And then prior to that, I had consulting roles, know, on Cineon, Lloyd, et cetera. I've done my rounds with the big fours, advised larger clients, lived through some interesting war rooms. And we are back again here, you know, 2026. And again, people are scared again, once in their life again, this is... almost the industrial revolution coming back in a very different way. So I am excited. I think there is the fear is genuine and every forum that I speak, people ask, what should we do? And my thought process is very straightforward, right? It's either TheUnpackshow: Yeah. Harveer Singh: we adapt or we perish as simple as that. And that's what as humans we've done. We've always adapted and we will always adapt. It's just TheUnpackshow: Yeah. Harveer Singh: a matter of showing the path and leaders have to step up. So that's a brief background. Currently, I'm a co-founder for RIS wireless and RIS network. We are a new age MVNO. We have an AI enabled platform that essentially offsets the cost of wireless of the customers. It's a new age concept where think about we're bringing the American Express experience to wireless where your $8.95 plan gives you amazing rewards. We do the same thing on the wireless side, but bring it to the common man. here. We're getting some great receptions. And again, because we are nimble, we are small, our goal is to be very customer TheUnpackshow: Yeah. Harveer Singh: centric. And the AI technology is helping us give back to other customers the rewards that they actually deserve. again, hands on, building it, doing it, and can share what the learnings are as well. TheUnpackshow: Yeah. Couple of things, Harvey. Like one, yes, I read the When Data Moves book. thanks for putting together you know, the way you told the story in the book is fantastic. I read the book. Thanks for that. And then second thing, you know, for those who are who are interested in what Harvey is doing right now, I would definitely recommend checking his website. Harvey, if you can tell the website name, you know, it's interesting, you know, just go look at the website, right? It's completely different. It's not a normal, useful website, so go and look at the website. Harveer Singh: So resviolist.com, we had imagined that in a world where there is so much lack of attention these days, there's so much crowd out there, how do you stand out? So our team decided to build art in a way which tells TheUnpackshow: Yeah. Harveer Singh: a story. Every art tells a story in the traps that we are. And our designers are ex-Pepsi. They broke that call it they broke the internet when they came out with their ads back in like the early 2000s. So they've done it again. So very proud of that. And then our essentials platform, which is very unique, which is the offsetting platform. We are just about to launch the AI powered rewards in the next few weeks. So stay connected and stay put for that. We will be making some big announcements there. So excited. TheUnpackshow: Yeah, yeah. Thanks for that. It's a unique website. I want I just want to highlight that highlight that on my show. Harveer Singh: Thank you. TheUnpackshow: so my first probably like you know discussion point for you, Harvey, is like you when organization launch mini AI initiatives, right? The initial conversation usually focuses on technology, choosing the right models, choosing the right platforms. Choosing the right infrastructure, right now, and also making sure that you know what kind of guardrails that you can put, right? So those are important decisions that you need to make when you launch an AI initiative. But however, you know, we've been learning that, you know, it's consistently showing up, right? You know, the technology alone rarely determines the success or failure. Most importantly, the often struggle, the organization often struggle, not because of the technology, right? because but but but because the employees, right? Don't fully understand how AI fits into the, into the into their work, how it fits into the broader organization, how it's going to differentiate, how it's going to differ, right? You know, so all those c questions are popping up. Some are like smart enough to adapt, to adapt and use AI on a daily basis, right? But my question to you is like you have worked in organization, especially in the banking, and you're doing some cool work, you know, with with your founding company. Why do you believe, right? People, not the technology are the most important factor in successful AI transformation. Harveer Singh: Absolutely. I go back to this example when I was in, I think the 2008 when the whole crash happened, when the emphasis on data became more important. One of the examples that I've used with my clients then and that I use with my employees right now is that, look, right now we are getting into a phase where you have a new technology that has come on board, essentially automating TheUnpackshow: Mm-hmm. Harveer Singh: sub automating certain tasks, making certain decisions on your behalf and doing some tasks. Is this new? No, not really. Your car, TheUnpackshow: Yes. No. Mm-hmm. Harveer Singh: car, car is doing some of that already. Right. And it's been doing it since the last 50 years. Right. You, you, you drive the steering and the car moves. You don't go move the tire. Right. There is an automation that is built in and then Tesla took it and said, okay, we will make that decision based on you know, the road conditions, et cetera, et cetera. So that technology is already available. Now, the job of the car is essentially to transport people. The same way the job of an organization, let's say a bank, is to provide financial services to its customers. It can choose many different vehicles to do that, which means that, you know, do they need a Ferrari to provide a service? Probably not. Or probably yes, depending on the class of people that they might be servicing. But on an average bank, I think every bank has the right technology that is already there, that they don't need to replace most of it. There is some modernization obviously required, there's wear and tear happening, there is stuff happening, but it's already there if they start to think through how to bring it all together. And it all comes down to how do you connect your data? How do you connect your underlying pipeline? How do you make that available? And then on this, and how do you manage your stakeholders to align onto the common goal? Majority of the times, what I've understood at what I've seen is you talk to different parts of the business, they have different goals. They do not resonate with each other as well. AI has to fall into TheUnpackshow: Mm-hmm. Harveer Singh: the same trap, right? You have one part of the AI organization saying that I want to... maximize the revenue. The other part is going to say, I want to minimize the burn. Look, TheUnpackshow: Yeah. Harveer Singh: they can be, these are very opposing intents. And if they don't take those two intents together and build a cohesive element of that strategy, then neither of them will succeed because one will keep cutting the cost. The other one will keep trying to build on the cost to bring more revenue. But at the end of the day, no one will succeed. So It is a human problem to align on the aspect of what exactly do you want this AI to do. Majority of the time that is not clear. Now, TheUnpackshow: Mm-hmm. Harveer Singh: in one of my previous, I just visited a CTO club initiative here in Dallas that I'm doing with another CTO friend of mine. you know, we had a bunch of CTOs and they asked him the same question is like, Hey, have you guys seen any success with AI? And my first question is defined success. If you TheUnpackshow: Mm-hmm. Harveer Singh: mean that your workforce is now grammatically correct on every single email, sure, it's success. But has it really produced any output for you? Probably not, right? So the measurement of that success, the KPI of that success is also not defined yet because we are in such an exploratory phase in our lives. The big promise that sits on Instagram is that now you can build a billion dollar company. Now you can do XYZ. Now you can do Y. I am the one who's doing it. I know exactly what the limitations are. And I can tell you for a fact, it is not that easy for everyone to do it. And especially when you are in a regulated environment like a bank or a regulated environment like a healthcare industry, just letting AI lose on something, it's not going to make a whole lot of difference. It's actually going to be detrimental to what you're doing. So it is the... accountability of the leadership to shepherd the way. AI can perform a task. It is never going to be accountable for that task. It could potentially be responsible for the task, but not accountable. So having a very strong RACI model now in the engendic world and the AI world is the need of the hour. And I don't think most of the organizations have built it yet. They are all into that, fine. know, what tasks, what, processes we can automate, what is the answer that we can get, but I think they need to step back and say, okay, what accountability that we have to put on that first, before we start actually experimenting with more. That would be, I would say that is how the leadership should start defining and adopting it. TheUnpackshow: Right. Yeah, I think I think you're making a good point, Harvey, right? So when we talk about people and leadership, right? So I think we need to talk about the culture as well, right? So what does an AI ready culture will look like, right? For example, you can buy the technology, but you know, you cannot buy the culture, right? So organizations that successfully embrace AI, you know, tend to have certain cultural characteristics characteristics, if you will, right? so I think they encourage experimentation instead of perfection, all those good stuffs are there, right? You know, they they also reward learning rather than simply rewarding the expertise side. Those things are there, and also they in every organization is trying to create this psychological safety so that the employees feel empowered and comfortable asking questions. All those good stuffs are there. But when it comes to AI ready culture, right, it's not just built through a single training session. it evolves through leadership behaviors, organizational values, and everyday habits, right? Because you're using AI for every single thing nowadays, right? So, in your view, right, as you as you are embracing AI, you know, with your founding company and also embraced AI in your past, what does an AI culture or AI ready culture actually look like in practice, right? Forget about the theoretical behaviors, right? what are the most importantly, what are the behaviors, right, and the leadership actions that everyone should take, right, to create that kind of an environment. Harveer Singh: So this is probably a very interesting question. And I'm going to answer it in three ways. One is any type of change, when it happens, there's always going to be three types of people or three types of results come out of it. There will always be people which are resistant to change that are going to pose roadblocks or potentially because they think it's a threat, it's always going to be there. So that's human nature. Then there will be people who will just do enough to get by. And then there will be people that will accelerate and move the company forward in these change initiatives. Now, I always say that all three are required. The bottom ones which are resistant to change. They are resistant to change because they know the processes very well and they have an understanding that these processes could cannot or could not change. That is not a wrong assumption. That is just how they feel and how they fear. Right. So that's one. They could be your biggest critic and they could be the ones that could lead you down the right path when you are deviating. So treating them as threat and treating them as they are not ready to change is the wrong way of dealing with. The ones that are in the middle are the most, potentially will be the most inefficient, which is just get by because they know a few things here and there. I think the top ones, if they join forces with the really bottom ones, which means that now they have expertise. Look, AI can perform a task. I mean, I personally build an application to manage some of my affairs on a weekend. Is it production ready? Probably not, but is it good enough for absolutely. It's great. My family uses it. You know, my son uses it. We keep, keep things on schedule. I don't have to buy an app where I can, you have to manage my expenses. know exactly what my expenses are. I'm monitoring it. I cut down on things myself. Why did I do it? Because I can. Right. I haven't quoted in a very long time, but now, you know, I always call it, you know, these augmentation tools as like, you know, Red Bull of the world. They have giving you wings. Now you have to fly. You have given wings, but you have to fly. And at the same thing, when I was at Truist, I had met with so many people who have been with the organization for 20 plus years, know, maybe 30 years, maybe 35 years. And when you talk to them about processes, their eyes light up, right? They are absolutely gold because they know where things are hidden. where the skeletons are buried. So from a cultural perspective, an AI ready culture is which will enable these people to come out of those so-called closets and embrace them without showing them that they are aged or they are redundant or anything. They are not. They are going to be your best fight against competition because they know these processes well. And it is well proven. any model, give it to two different people, the output is completely going to be different because of the way you prompt, the way you ask the question, the way you describe the problem and the way you TheUnpackshow: Mm-hmm. Harveer Singh: interpret the processes. So if I was to put someone who does not know, who only knows AI into a, let's say, a customer onboarding process and a person who is manually onboarded customers for years and he knows everything that needs to happen. onto the same process, the output of the second person will always surpass the first person, it is guaranteed. So somehow, there has to be this trust that has to be established with the workforce that your expertise are the fuel that is required to power this AI. In my opinion, there are three fuels, data, the people, and then the processes. Technology is something that is going to be so readily available that TheUnpackshow: Mm-hmm. Harveer Singh: you can swap things in and out very easily. can take any model that you need. Every cloud company will offer the same kind of models that you can pull on. That is not going to make the difference. It's that three things, TheUnpackshow: Yeah. Yeah. Harveer Singh: three fuel, the people, the process, and the data will make that culture really, really good. And acceptance that... we are either ready or we are not ready is going to make a whole lot difference. Now, this week, I released a skill on Cloud personally. And the answer for this was exactly what you asked me. I released a skill called Data Pulse, which essentially enables an executive to run that skill in their local environment. It's a markdown file. It describes exactly how to measure the pulse, the data pulse of an organization for AI readiness. And I encourage people who are on the call, you know, look me up on GitHub. It's my github slash harveer10x. Go download this tool. Learn exactly how to measure the pulse of your organization because behavior, personality processes is partly built into that skill itself. Do it yourself. You will understand where you lack and how to become that. that AI ready for your organization. TheUnpackshow: Right. Yeah, well said, Darweer. I think, you know, we talk about human intelligence and artificial intelligence, right? So a lot of debate on human in the loop versus human on the loop, right? Is it going to eliminate humans? Right. So I think I think well said. you know, my my next segue is like, you know, how the next decade, right, we'll likely we'll likely bring AA agents. I think, you know, you mentioned you have built your agents for your personal stuff and thanks for sharing that link, right? so I'm sure everyone is having going to have an AI agent. I I'm I bet like, you know, the next podcast that I'm going to do is going to be my agent talking to your agent, right? That's going to be the the the scenario. so but but but most importantly, right? You know, colla you know, AI agents are capable of collaborating with humans, making recommendations, coordinating workflows, all those fun stuffs are there, right? I think the jobs are going to evolve. The winners won't simply have the smartest AI, right? They will have the most adaptable people and also the strongest culture, right? That we just talked about. But I'm curious, you know, based on your learnings from the peers and based on your learnings from the leaders, if you look ahead, like maybe not five to ten years, maybe two to three years, right? What what is exciting for you? What is the most exciting stuff for you, right, in the future of work? and what advice you would provide or you would give the leaders and also the professionals, right? Especially the eng professionals for their next chapter in their life. Harveer Singh: Yeah. Again, I'll take two, two different variations of this question because, you know, I'm dealing with the people who read my book and talk to me about data, human spirituality. Then I have my own work, like, you know, on the, on the wireless, wireless side and the residential side, where we are actually experimenting with this technology itself. What you talked about agent talking to an agent, because we are underlying a cellular company. are building a technology, which basically enables an AI binding to a cell phone number itself, which enables them to talk to each other redundantly over a cell network. If the data is not available, they can talk about talk through SMS. They can talk through just radio frequency as well. we have a provisional pattern that we have just received on certain aspects of binding. So that's pretty interesting. So one thing is that if you, the next three years, There is no doubt this is going to be 10x or 100, even 100x in terms of the uses of this technology. There is no doubt about it. You will see that it's embedded into pretty much every single edge hardware that you will end up procuring. My challenge always will be that how much you want to trust this technology to autonomously do this and if you want to. How much of this is you want to give access to your life? And what is your contingency if there's something that doesn't go this way? We are actively designing a system where, you know, and this is interesting on the wireless side, we are saying that if you bind an agent to a cell phone number, which is a prepaid number, I can cut it off whenever I want to. And then, you know, life goes on. doesn't, it saves your privacy and it protects your privacy. So we are already thinking and we are already going down that path. But people also have to start thinking how much to give access, what to give access, where to give access, because, know, abuse is again, you know, is, it's, it's, will happen, right? So it will definitely happen. But what I'm really excited about is that there are so many tasks that I personally have in my mind, not tasks, but ideas that TheUnpackshow: Yeah. Harveer Singh: I have in my mind, that if I was to experiment with all of them at the same time, I would need an army of people. So somehow, you know, with how I or somebody like in my position is thinking, they don't need to waste their team's time in ideating stuff. They can ideate themselves. And this will be a biggest difference between a good leader, a great leader, and an exceptional leader. Like if you're in a leadership position, you're already good. No doubt about it. The organization did a good job choosing you. Great. Congratulations. But if you want to be an exceptional leader, you need to understand how to set the priorities right for your team. And that you can do so much of that vetting beforehand, which aligns with the goals of the organization, which aligns with the goals of your customers. Because majority of the time, you know, we don't do that. We are always under pressure. There are so many projects that are going on. It is hard to understand how to prioritize that in that environment, in that high pressure environment. a leader can become a product person, an architect, a developer himself. And he doesn't need too many people on his side. He can just do it himself before he goes to his team as a guys. I think this is the right way to go. And that will eliminate so much rework that the team sometimes have to go through. I call myself the culprit sometimes, you I have an idea and I was like, Hey, team, let's go build this. And they've gone through and build this and I'm like, you know what? doesn't look good. Yeah, well, we'll change this. I can now go back and say, what could I have done differently? mean, 10 years, five years ago, I didn't have this, this technology. We had these dashboards. Every time you build a dashboard, like, no, this is not doing the right thing. You know, let's go to bed. TheUnpackshow: Exactly. Harveer Singh: So I think it, it, as a leader, this could be your greatest asset. The second thing in the next two to three years is what I expect is. these models to become more and more accurate. Right now, the problem that I see is the hallucinations that come in these models. You can make certain decisions, but you can't make accurate business decisions based on the model itself. Because, you know, let's give you an example, right? You ask the AI, what is your bank account? It says, yeah, it's maybe like $100. Do you like the maybe, or you want it is $100, or is it $5,000, right? You cannot do a maybe. And then you go ask them, is it a hundred dollars? No, no, I just assumed it based on like, know, 12 years summary. I'm like, no, that's not what I asked. I have a real example I'm sharing. This is common in every single household. My wife wanted to go shopping to an outlet mall. And she asked me for a, is that shopping mall have a particular store? Now TheUnpackshow: Mm. Harveer Singh: I said, okay, I'll go look it up. She's like, no, no, no, I'll find, I'll just ask Chad GPT. And I'm like, Well, chat GPT is not a search engine. She's like, no, it'll answer. She went down this path and it answered and it says, yeah, they have the store, et cetera. said, trust me, that store is not there. no, chat GPT says it. I'm like, no, it does not. It's not there. So I showed how to go to the website. I said, look, here's the website. Let's go through this. We didn't find it. Then I said, write this back to chat GPT, challenge it and say this store does not exist. And chat GPT responded, you you're right. I was assuming it because this brand, this big brand, the outlet brand, like the malls normally carry this brand. I'm like, what if she started driving like two hours away from Texas and in the middle of it, she wanted to return something to that specific store and it's not there now two hours down the line. And this is the problem, right? One, she's not trained to ask the right question. mean, she asked the question, right? But this is so basic that, you TheUnpackshow: That's not good. Harveer Singh: It's a, and it's a real life example. This is exactly what happened. And it, it would have left. It would have led her a day down the drain with gas money gone down the drain. And if you take this an example and expand it to an enterprise, think about what can happen, you know, and I collect these stories because I'm going to write another book around it. Right. And this, this is a story where I told my wife, I'm like, look, I'm going to make TheUnpackshow: When AI moves. Harveer Singh: you famous. She's like, yeah, you're making me look. bad. I'm like, no, but this is a common thing. I do that mistake all the time that I ask a question and five prompts later, I'm like, you know what? I should have asked this prompt first. And I don't think there is enough education. There is enough knowledge around what people are using. So that's why the token burns are higher. TheUnpackshow: What? Harveer Singh: So from a productivity perspective, I think people will start. So the same way we do Jira stories and Jira points today, which we say, okay, know, a task takes X amount of time to complete. This is your cost that is required. I think we will be adding the number of tokens that is required to complete this as well. And essentially that additionally becomes your ROI metric, right? Again, my two weeks ago, I launched this again, a skill called earned versus burnt metrics, which means that how much value have you earned? And how much tokens have you burned? So ours is one, but the tokens is another one. can, you can burn 2 million tokens and, and you say, this is good enough. No, is it production ready? No. Then you have burned, you have not earned. And encourage TheUnpackshow: Yeah. You're on yeah. Harveer Singh: people to, think about what do you have you earned and what have you burned? You know, check out that scale again. It's I, I, I've been publishing this open source because I know people will require it. The faster you adapt to this in your Jira stories, the better performance and the better productivity you'll get. And I'm hoping these will become more like a norm than a deception in the future. TheUnpackshow: Yeah. Yeah. Thanks sharing that metric. Earned versus burned. And thanks for sharing your story. I stopped using search, Google search. You know, this is an interesting one. You know, I I I I didn't think of those, so probably I'll play with that. And and and thanks for sharing the the the practical examples, Harvey, right? So and and thanks for being part of the show, Harvey. good to talk to you and I can keep on going, but I just want to limit the conversation here. But I enjoy the conversation, a lot of good insights. thank you so much, Harvey, for your time. Harveer Singh: I appreciate it. Thank you. Take care. TheUnpackshow: Yeah.