Sanjay Srivastava - Chief Digital Officer at Genpact

(Joel Beasley at 00:00:00) Hello, my friends. Today we are talking to Sanjay, the Chief Digital Officer at Genpact, and we discuss the four industry trends that are powering digital transformation, how ethics will play a role in the advancement of artificial intelligence, and what the future of work will look like as more processes become automated. All of this right here, right now on the Modern CTO Podcast. Here we go. This is the Modern CTO Podcast.

(Joel Beasley at 00:00:37) Are you professional? Have you ever been paid to model before?

(Sanjay at 00:00:42) I've been—many people have helped me set up all the lighting, so thanks to them.

(Joel Beasley at 00:00:47) It looks really good. Got a lot of books back there.

(Sanjay at 00:00:51) Yep.

(Joel Beasley at 00:00:52) You read a lot?

(Sanjay at 00:00:54) I used to. The hard copy books. Now I read everything electronically, but I used to read hard copy books. So many of these are from business school and engineering school and then obviously all of the technical books. So yeah, I do read a lot.

(Joel Beasley at 00:01:09) What I do is I do the Audible. And then when I really like the book and I want to remember it, I buy the physical copy and then put it up in my office. So when I walk by it, I see it and I think of the principles in the book.

(Sanjay at 00:01:21) That's a good way of doing it. Yeah. I guess the audio thing works if you work out or if you drive because I know people that do that, and they or they run. It's an awesome thing to do. But if you don't—I don't do long commutes, or none of us do now in our new work-from-home environment anyway. So then it loses its thing.

(Joel Beasley at 00:01:39) So last time—and so this is the second time we're talking, but it's the first time the episode's going to air because last time we had some audio quality issues. So I'm very sorry about that.

(Sanjay at 00:01:49) Happens once in a while. I'm sorry that it happened on my end, but glad that we could do it again.

(Joel Beasley at 00:01:54) When I heard that, I was like, oh my goodness. I was like, we have to get Sanjay's brilliant insight and advice out to the world. This is unacceptable. So I'm so glad we're able to do it again.

(Sanjay at 00:02:05) Oh, you're very kind.

(Joel Beasley at 00:02:07) So we're just going to hang out and talk. Is that cool?

(Sanjay at 00:02:09) Yeah. Absolutely.

(Joel Beasley at 00:02:10) Awesome. So one of the topics I was pretty excited to talk with you about was this content I saw that you put out for industry trends for digital transformation. So I always try to find out where the areas that people are experts in and let them share so we can bring the best knowledge out to the people. So I'm curious, what's going on? What are the trends for digital transformation?

(Sanjay at 00:02:35) I think, Joel, one of the—before we get into the trends, one of the largest things that's happening is that digital transformation has become a boardroom imperative for most corporations. As we get through the pandemic and as we start reconfiguring our business models to the new normal and then start thinking about how do we grow out of the pandemic, the reality is that digital transformation becomes the path and the way to do that, and that's becoming pretty high on everyone's list. And as a result of that, and you've seen different numbers—you know, two years of digital transformation is getting there in two months. I saw in one place four years of digital transformation in four months, et cetera. Right? But I think if you go one layer deeper, what's happening behind that large-scale drive are four trends that are coming through.

(Sanjay at 00:03:09) The first one is cloud journeys are accelerating. Now cloud was always known to be a great economically viable alternative. It was seen to be a place you could manage and control your infrastructure centrally. But what's happened is the pandemic has taught us that cloud can scale, and it's actually answered one of the biggest questions that was on paper, which is, can the cloud actually scale? We've answered that question. We've had a lot more people, some reluctantly, now become users of the cloud. I mean, the technology you and I are using today is a cloud-based technology just to talk. And so more people have sort of bought into and understand and personally experienced the benefits of it. And then the reality is that as you try and bring in new capabilities to address the gaps that have come through because of pandemic, you need a new technology back then. You need a business technology architecture that gives you those new capabilities. And really, the only way to deliver them at speed is on the cloud because, and using the cloud as a backplane. So we think cloud journey is accelerating. That's trend number one.

(Sanjay at 00:04:04) And the second big thing that's going on is the role of data and analytics. Data and analytics were always important, but now each of us are relating to it in a very different way. I mean, my parents, for instance, who were planning to do some travel, they become data science experts. They wake up every morning. They look at papers. They understand flattening the curve. They understand graphs. I mean, they're data scientists trying to figure out what's the right time for them to get on this flight as an example. And if that's the case in our personal lives, can you imagine what's happening in the enterprise? Business leaders are running a buzz around figuring out trends and decisions around supply chain and dynamic allocation and parts inventory and e-commerce trends and so forth. So that's the second big thing that's happening.

(Sanjay at 00:04:58) Then the third one is this end-to-end digitization. The reality is if you look at any business process, the front end always got the investment to get digitized. So we see all these e-commerce front ends. We can go to portals. We can shop electronically. Right? And all of that happens in the front end of the entire value chain of buying something, manufacturing it, shipping it, and getting paid for it. Right? But as you start walking back into the rest of the ecosystem, the rest of those capabilities hadn't really fully digitized. And so you had this kind of balance between front end that was there and the middle and back office that wasn't quite there. And what's happened is in the pandemic, many large enterprises have gone from things like 50/50 online retail. So the percent of your goods that you sell one way or the other. They've gone from 50/50 to 100, and it's happened overnight. And so you've got all of this commerce coming through on the front end, and then as it hits the middle office, it hits bottlenecks. And so the supply chain becomes an issue. Allocations become an issue. SKUs and just the proliferation of SKUs become an issue. And so you kind of move on through all of that. And what's required is to digitize it end to end. And so this end-to-end digitization that allows you to kind of give the same capability across the entire customer experience becomes very important. So those are three big trends.

(Sanjay at 00:06:08) Look. The fourth one is we're all adopting AI, artificial intelligence, more and more. And artificial intelligence got a lot of positives, and I'm a big proponent and believer. We do so many projects in artificial intelligence. But one of the things that always gets sort of missed in the loop is the role of the human and the human in the loop. And I think whether it's sort of, you know, thinking about the human in terms of designing better AI, thinking about the role of the human actually curating and tuning the AI to be more higher accuracy, or in being able to use the AI to drive higher level predictability and make better judgment calls, that role cannot be forgotten in all of this technology discussion. And so we're seeing a big trend line around thinking about human in the loop and proactively planning for that. So, Joel, those are the four big things I see in the industry today.

(Joel Beasley at 00:06:56) Oh, amazing. Yes. You did that so well. I'm sitting here and I'm like, you're absolutely crushing it.

(Sanjay at 00:07:03) Well, I'm living it every day, buddy.

(Joel Beasley at 00:07:05) Yes. That's my next question. So when you come up with these points, do you—you are a speaker. You're an executive at the organization. You're always doing this. But do you set aside time and actually write these points out and collect your thoughts? How do you get them so refined and down?

(Sanjay at 00:07:30) Well, look. I mean, I don't think I have a perfect answer for you, but I think it's a combination of a few things. And by the way, I'd probably give you a different answer today than I would give you six months ago. So in the last six months, I've been on zero business flights. Maybe not six months, in the last—I guess since February. Right? So I haven't traveled at all. All my interactions are through video with advisors, with clients, with peers, and obviously with our employees and our teams. Right? And so that has its own set of challenges because the fatigue on Zoom—there is such a thing as video fatigue—is really high. You know, we're not really working from home. We're actually living at work. We're no longer working from home. I mean, I am actually living at work. It's just that work happens to be in my home. And so that introduces a whole set of challenges around how do you continue to be creative? How do you continue to engage in collaborative discussions? How do you think about weak ties? Right? It's very easy to maintain your strong ties, the people you obviously always network with and you speak with and you directly interact with. But how do you go after your weak ties? The people you won't actually normally talk to, but you might bump into if you went to the office, if you did travel somewhere, you went to another city as an example. But I think a lot more thinking has to go in by all of us and how do we manage and maintain our weak ties because those ties contribute a lot to our intellectual thinking. So anyway, so I think things have changed quite a bit.

(Sanjay at 00:09:02) But I think to your point, I don't think I do it well, but I think the things that do work for me—number one is spending a lot of time with users, with implementers, with sort of the digital-minded folks across the globe. So you need to find more ways to have those entertaining and exchange of ideas. And then I think a little bit of it is being structured and translating some of the things you're hearing into a few takeaways that, for instance, my team has to take and then convert these high-level ideas into actual things that will deliver. Practices we build, frameworks we develop, IP we put together and bring it to play. And the reality is there's so much going on in digital. It's a problem of plenty. So here's my advice to every single board I meet with. Pick a few things because there's way too many things to be done right now. Pick a few things and then you get through them. But how do you pick the right ones? And that's got to be based on a broad base of listening, high levels of curiosity, good amount of reflection, and then iteratively collaborating it and testing it with lots and lots of people.

(Joel Beasley at 00:09:51) Is that something that Genpact does as a service? Is that something you guys do as part of your business?

(Sanjay at 00:09:57) Look. So just 30 seconds on Genpact. We're a transformation company. We help large corporations. We serve Fortune 500 companies mostly. We'll take those companies and help them transform from who they are today to who they want to be. And a little bit of that is not just branding and marketing. It's actually about the core business processes. How do you go up to new markets? How do you think about new products? And then how do you deliver new experience to customers? Now part of that is you have to fix your supply chain. Part of that is you have to think about your finance and accounting. Part of that is you have to think about how do you do sourcing and procurement. Part of that is how do you manage regulators, if you're in banking or some of the regulated industries like pharmaceuticals. How do you manage all of that stuff to sort of keep up? And the ways we all used to do it was set up really well for the business we were in and was mostly manual. And even if it was automated, it required automation and then people looking at it and making judgment calls. But technology has shifted, and available toolkits have gone up. And so there's an opportunity to take that and reimagine those business processes with the use of new capabilities and analytics, and artificial intelligence, and automation, and experience, and put that into place so that new digitally transformed business process looks and feels different, but most importantly delivers results.

(Sanjay at 00:11:11) And so the 30-second answer to your question is, as a company, we do three things. We bring digital technologies to play, which obviously I spend a lot of time on, but we also bring a lot of domain knowledge and process insights because we think that's integral to actually orchestrating the technology in a meaningful fashion to drive results. And then the third thing we do is, particularly for large corporations, you need a programmatic execution of people, process, data, and digital, almost four different things. And you've got to sort of synchronize the execution of those things so altogether they end up in the right place. Because if they don't, you don't get the return on the investment. And so we have this notion of transformation services, this programmatic execution of four different components in synchrony to get to an endpoint and a business outcome. And that's what we do.

(Joel Beasley at 00:12:00) I was taking notes. It's people, process, data, and...

(Sanjay at 00:12:04) Digital. Technology. Digital.

(Joel Beasley at 00:12:06) Okay. See? This is very useful conversation for me.

(Sanjay at 00:12:09) Good.

(Joel Beasley at 00:12:11) So you're helping these companies do that. And you talked about some of the trends in AI. Have you gotten to see that Netflix documentary, The Social Dilemma?

(Sanjay at 00:12:21) I did see that. Yes. The—it's a 19-minute or 17, 18-minute quick look. Yes.

(Joel Beasley at 00:12:27) So there's a longer version. You might have seen a preview that Netflix did, but I'm sure you got the gist of it. But I was curious about this concept of algorithms and echo chambers, and if there's any interesting solutions out there or if you have any interesting thoughts on basically the social media algorithms—they're optimizing to get you to spend as much time. The outcome they want is you spending as much time as possible in the platform. So they show you content that will get that to happen. And I was just curious what your thoughts are on that in general.

(Sanjay at 00:13:02) Look. It's a really interesting subject because the one side of it is it's really bad and it's really evil, and we're manipulating people to the wrong outcomes. Right? I also have a view that looks at the other side. You have to balance these things. The other side of the thing is we used to do a bunch of things manually. As human beings, you and I are talking. You're actually looking at me and you're looking with your eyes at me, and that's drawing me closer to you. And I'm able to communicate and converse with you in a different way than if you were looking off in the horizon or if we didn't have this video. Right? These are things we do naturally. As human beings, you and I do it together. I'm moving my hands around because I'm trying to emphasize different points. I'm not even thinking about it. It's just happening. But what did I do? I just, quote unquote, manipulated you to listen to my sentence and pay attention to the two words that I thought were important. Right? So the flip side of that discussion is that, you know, you've got to look at it both sides. The flip side is it's things that we would do anyway. We now have the ability to amplify it, to be able to do it at scale. And so you've got this amazing toolkit that has all of this wondrous possibility. And I think the trick comes down to balance. It comes down to how do you manage, you know, how do you manage the line between stepping too far and doing enough. Right?

(Sanjay at 00:14:20) And it's a thin line, and it changes by people, and you'll have a different perspective than I will have. One of the things I get asked a lot by boards of large corporations we work with, and I always have a point of view on this, which is I actually think we need digital ethics officers. I think boards of large corporations—you know, it used to be twenty years ago there were boards, and then you had audit subcommittees, and you had a compensation subcommittee. Right? I think ten years from now, you'll have ethics subcommittees, and I think large corporations are gonna start paying a lot more focus on this.

(Sanjay at 00:14:51) It's just a requirement. And these ethics officers, these digital ethics officers are gonna have to make these hard calls, the trade-off between where the line is and how far do you wanna go before you go over too much. And I think the documentary you talked about actually does a good job of exposing one side of the equation, and I think you have to balance it with the other side and come to a medium. And, you know, the medium you come to is gonna be different for you than for me, but each of us have to take a point of view.

(Joel Beasley at 00:15:16) Yeah. When I was—

(Sanjay at 00:15:16) That make sense?

(Joel Beasley at 00:15:17) Oh, a hundred percent. When I was watching it, I was like, the first ten, fifteen minutes, I was like, alright. Maybe I just turn this off because this is so clearly just slamming everything. But then I thought, okay, well, this is a different view. Let's just watch the whole thing and balance it out like I do with everything else, with all the other digital diet I have. Right? And so I watched the whole thing and I said, okay, this is interesting. But one of the things that it leaves out or I didn't think that it addressed properly was the concept of personal responsibility and individual ownership.

(Joel Beasley at 00:15:51) Like, all of those things they mentioned only take place when I'm choosing to go to these places. It's like I'm choosing—like, if they were buildings that we would walk into, it would be a much different thing. Right? It's like I went to that restaurant or I went to that bar and then this event happened or when I'm inside that environment, I'm aware of what's going on or aware that there are things going on that I'm not aware of. Right?

(Joel Beasley at 00:16:18) But I'm still choosing to go in there. And I understand that there's a whole study of addiction and all of that, but we're still doing it. Like, we're still going there, and that's just the landscape the way it is today. And me personally, I love the fact that there's all these new operating system features specifically in Apple, just because I happen to have the iPhone where they're forcing you. They show you the screen time that you're using. They give you options to limit it between yourself and your kids. So, I mean, I think we're definitely headed in the right direction. We're doing what we can. There's just not really clear answers to some of these things. And I think getting us on the path to getting clear answers would be useful.

(Sanjay at 00:17:10) Yeah. And I think you hit the nail on the head. I think in many ways, it's a personal decision. And I think what these kinds of things bring up is the need and the necessity for us to reflect as individuals. What's the right thing for me, for my family? And by the way, that's gonna be different for you. I can't expect a government policy to decide how I want it to be for my family. And so I think what it does is it raises it to the level of consciousness. It's some of these things we need to think through. And then I think it forces us to reflect on it and make the right decisions. And I think that personal responsibility is a big part of that discussion.

(Joel Beasley at 00:17:44) Yeah. I mean, this is something like the founding principles of the nation that we choose to live in. You know? So, like—

(Sanjay at 00:17:51) That's right.

(Joel Beasley at 00:17:52) I've actually gotten interested in the concept of free speech lately and just out of the curiosity of, you know, what it means and the origins and the differences between countries and their free speech policies. And I didn't realize that that's one of the things that makes America the greatest place ever is because of the free speech.

(Sanjay at 00:18:15) That's right.

(Joel Beasley at 00:18:16) I'm a huge fan of that. But we'll get back on topic with some CTO stuff.

(Sanjay at 00:18:21) Okay.

(Joel Beasley at 00:18:22) But, actually, I'm curious about your—I noticed in your profile that you do a lot of work with startups. You've been on advisory boards and a part of accelerators and things like that. Are you seeing any really cool, interesting technology come about in the startup community right now?

(Sanjay at 00:18:40) Tons of it. Yeah. That's one of the reasons I love being part of startups. By the way, my background also was with startups. I've built four startup companies all in the Silicon Valley. Each ended up being good pieces of technology that are part of a lot of corporations now. My last company was acquired by Genpact, which is how I came to be here as an example. So I've always loved the innovation, the creativity, the agility, the extreme focus to building something that addresses new gaps. I think it's been fantastic. And I, as you picked up, I continue to spend a lot of time with startups, mostly in advisory boards, and through incubators and venture funds.

(Sanjay at 00:19:19) But to your question, I think there are a number of things that are coming through. Now I don't look at everything. I look at things in a much narrower focus, but, obviously, outside of my area of focus, there's so much happening in the whole area of health and medicine, both in terms of how it's delivered, but also new medications and new proteins and so forth. So the whole bit there. Then outside of my area of focus is a bunch of stuff that's happening around physics. You know, how you get to smaller chips and smaller circuit boards and battery capacity and how do you get to a point where, for instance, one day we can fly airplanes on electric. Right? Which is—I mean, the future is all electric. That's where we're gonna end up. And so how do we fast forward the journey to there and get to Net Zero, etcetera?

(Sanjay at 00:20:00) But if you take that big spectrum of innovation and you reduce it down to the area I focus on, I spend a lot of time in artificial intelligence. I spend a lot of time in automation capabilities and analytics and new modeling techniques and bringing them to life. And then, significant time understanding experience, the role experience has in how you bring product and design out. And so in that sort of constrained set, if you will, not looking at the big thing—I think the intersection of artificial intelligence and practical business problems is a large area of investment. The reality is the world is gonna become more electric. Otherwise, the future is electric, so that's given. But the world is also gonna become more AI enabled. And as we start using artificial intelligence in our day-to-day lives, in our business applications, it pushes the frontier in a number of things. It allows us to do things that we traditionally thought were manual that can now be done by artificial intelligence, and then allows people like you and I to move to the next value up.

(Sanjay at 00:20:59) And I think there's a lot of work that needs to happen around reinvigorating and reskilling our employees and our kind of global working population to take advantage and not get displaced by, but take advantage and enhance their lives through the use of artificial intelligence. So in AI, particularly around computer vision and physics-constrained applications like manufacturing supply chain, predictive capabilities around machine failure, parts replacement, those sorts of things. There's a ton of new stuff going on around artificial intelligence and business process. You know, our definition of business process was we put in an ERP, an enterprise resource planning system, and we were automated. Well, what did an ERP do? It automated some of the functions. What happened to the rest of it? Well, you now would get on and we'd enter some information. We'd look at it. We'd make a decision. We'd call someone. We'd follow up on an email. We'd check back in. And we thought the whole thing was automated. Well, what was automated was this bit, which is what the ERP did, and then this other bit over here was very manual. We just did it. We didn't think about it that way, but that's what we did.

(Sanjay at 00:21:58) I think now we have an ability to look at that entire spectrum from what we used to automate to what we used to be doing manually and look at that again and apply artificial intelligence to actually automate all of that. That's bringing significant business benefits. It's not academic advancement. It's not a research project. It's not nice and cool things to do. It's real-world business applications that drive business results today. I think there's a significant amount of innovation that's happening at those crossroads. So, anyway, I'm very keen to see how that develops. I'm obviously participating in some of this. We use a number of those technologies here in my company, and then obviously advise a number of firms around it. So, yeah, lots going on.

(Joel Beasley at 00:22:39) Yeah. It's interesting. It's like as the technology is advancing, it's like we're getting this larger gap of things that we could do today and—actually, I think we're gonna need more entrepreneurs. Right? Because we're advancing technology at such a rate, and there's so much opportunity. Like, for me, I just look at the market and I say, take a step back and focus on doing more of what's working today because there's—it's so exciting. There are so many opportunities to bring technologies that are just out there and exist and just put them together with the right people and the right markets to create value. It's—I think we're at a better time than we've ever been before in the history of humanity with opportunity for business.

(Sanjay at 00:23:24) Well, I think you're a hundred percent right. I think there's a large opportunity that is emerging at a faster clip and pace than it's ever been before. But I actually think there are a couple other things also happening. The toolkits, you know, the bits and bytes that you bring to solve for the problem as these opportunities emerge—I mean, that has progressed significantly. And this is not even ten years ago, five years ago. Just in the last twelve months, you think about some of the developments that have happened around AI, and you look at the work that happened around OpenAI as an example or some of these new language processing tools, it's amazing. And so there's three things going on. There's large opportunities that are emerging on the back of all the transitions we're facing, the resilience through the pandemic, the growth requirements after that, etcetera, etcetera. There's a new set of toolkits, mostly using artificial intelligence and new ways of solving problems that are now on the table. They didn't really exist five years ago or ten years ago that you can now bring to bear. So that's number two.

(Sanjay at 00:24:14) I think the third thing that's happening is most of the low-hanging fruit is actually off the table. So what you need and to your point about entrepreneurs—you know, I think most opportunities now are complex. They're intertwined, and they need a convergence of different disciplines. So you need process understanding to combine with digital. Or you need to understand industry domains and combine with AI, or you need to understand the physics and AI, mathematics, at the same time to solve for predictive analysis. Right? And so this convergence idea, this idea of being able to combine two or three disciplines to collectively solve the problem in a different way requires a very different way of thinking. And that's why I think to your point about entrepreneurs, I'm a big believer in entrepreneurs. I used to be one at one point in time. I think they are key and central to this innovation spectrum that we're getting into because, you know, many of us—I spend most of my time with large enterprises. I'm caught up with the problems of today. I have to deliver return on investment for the people I work with today or maybe for tomorrow. I'm not necessarily getting enough time to think about five years out.

(Sanjay at 00:25:01) What entrepreneurs are able to do is to step out of that noise, and they're able to lift up a little bit. They're able to bring in collaborative teams that come from different disciplines. They're able to stay on top of new emerging trends. They weren't practically usable yesterday, but now have become practically usable today. And because they're thinking broader, they can bring that in. And then, of course, you combine those three ingredients with agile, with velocity, and then the ecosystems that have been set up. I used to live in Silicon Valley. I mean, that ecosystem was absolutely amazing from the point of view of all of the components you need pulled together to be able to bring a technology idea to market. All of that stuff is just available as a service. So you don't have to go reinvent the wheel on many of these things. And so it brings you a lot more focus on what you're trying to do.

(Sanjay at 00:26:05) So, anyway, I do think we're living through an amazing time because opportunities are galore. But actually now the toolkits have significantly improved, and this notion of convergence of different sciences or different domains or art, you know, that intersection is much better understood. And as we explore more of that, I think we'll achieve and be able to extract more value from them.

(Joel Beasley at 00:26:27) So with these entrepreneurs, because you get to coach them and interact with them, what are some of the common mistakes that you see happening with earlier-stage entrepreneurs?

(Sanjay at 00:26:39) Well, you know, I think there's two or three things that come to mind. Invariably, my first advice to entrepreneurs is solve the hardest problem first. So anytime you set off on a new initiative or a new venture, right, there's a set of things you have to solve for. Like, you gotta get this done. You gotta—this is—and it has to meet this requirement. It needs to be able to solve this, dah, dah, dah. And then you sort of, you know, one way to approach it and say, let's start with the smallest problem. It's just too much to be done, then we'll solve that, and we'll go to the next one. We'll go to the next one. We'll go to the next one. The reality is time is very precious, and you actually want to fail fast. Failing fast is not a bad thing. It's a great thing because it allows you to sift through ideas that aren't gonna make sense to the idea that will make sense much faster, and you can iterate to the right answer very quickly.

(Sanjay at 00:27:37) So number one advice, start with the hardest problem. If you're gonna go solve for something, right, break it down to its component problems. Pick the hardest one. Pick the one that you think is most likely gonna fail the most. Try and solve that first. Because if you solve it, then you work back down. If you don't solve it, you actually move on to the next idea. And I think this iteration is super integral to the process of innovation. So that's number one. I'm—the second one, and startups by nature do it better than large corporations. Right? Large corporations traditionally have a challenge with this. Startups have to focus, focus, focus. Right? You just don't have the investment capital. You don't have the mental bandwidth. You don't have the team capacity to solve for hundreds of things. So you gotta reduce the problem set to very small. And so you pick one thing, and you solve for it. And then you move to the second problem, and you solve for it. And that's a great thing to do because focus drives better results. Right?

(Sanjay at 00:28:28) One of my—if you ask me what is the hardest thing for me that I have to do, the hardest thing for me is that I have to say no to things. I have got more ideas, more opportunities, more great possible outcomes on the table than we can serve up, that we can deliver on, that we can take on. And so part of my job, by the way, is to say, yes, we will do that. But more importantly, here are the things we're not gonna do. Here are the things we will pass upon.

(Sanjay at 00:28:44) And actually have the conviction to say it. Because if you don't say it, if you don't put it out there, you leave that murky, teams around you, teams under you will get, they'll, someone will be running this direction, someone's going to be running in that direction. Right? And they'll be doing the right things because those opportunities exist, except that you'll be diluting it. You'll be spreading yourself too thin.

(Sanjay at 00:29:02) So I think the second advice is pick a few areas and go deep as opposed to pick a broad thing and try and do everything in there. I think that's the second big advice. And the third one is think about channels, think about customers, and think about adoption and the end users upfront. Design it upfront. So it isn't anymore about, I'm going to build a cool technology.

(Sanjay at 00:29:24) I'm not, I'm going to build an interesting product. I'm going to build something that delivers this performance that is way outside of what is currently delivered. That's not enough. That is not what it takes to build a successful business. You have to think about how you take it to market.

(Sanjay at 00:29:37) Who are the buyers going to be? Why will they buy? And then more importantly, how will users adopt it? With artificial intelligence, we're seeing results that are just amazing. But one of the challenges with artificial intelligence is adoption.

(Sanjay at 00:29:49) How do you answer the question of why should I trust this? It's a black box. I don't know how to, and you can keep showing all the results all day long. Right? And the results are fantastic and amazing. But till you can get people to believe in it and understand the logic or how it actually functions, you're not going to get the adoption. Design that early into your thinking. Think about it proactively and get that into your business plan. But they're the three things I would say to any startup CEO that I meet.

(Joel Beasley at 00:30:13) Man, this is great. Thank you. Thank you. Thank you. Yeah.

(Joel Beasley at 00:30:17) Like, one of my favorite things about doing this podcast is that I get to meet brilliant people and get to hear their takeaways from years of experience. And I love the fresh perspective that you bring. When I hear your ideas, and I'm trying to reconcile with mine and then also comparing them to past failures, and I'm like, oh, yep, yep.

(Joel Beasley at 00:30:42) Because I'm just always trying to figure out what's true and what works. But, yeah, that is actually great advice because figuring out who the buyers are and why they will buy, like, what's that point that they're so frustrated with they're willing to hand over cash. That is so important because there's a billion improvement features, but people don't like, they're just too small. They just don't have the attention or they don't have the focus or the timing because you kind of have to get this, I guess what I'm learning and I'm new to it obviously, but what I'm learning is the market has this motion and this movement and these trends and you really have to figure out, not necessarily a specific exact company at first, but what's happening in the market?

(Joel Beasley at 00:31:31) Why are people, where are they deploying money right now? And then how can I connect those types of people and then better validate the idea and the problem? And but it's interesting after you build a product and then, you know, you build all the features and then you realize the way you market it is not necessarily based on the details of the product. The way you're marketing it is how you're going to help the individual solve their problem. And that took me a while longer.

(Joel Beasley at 00:31:57) Yeah. A lot of pain there. A lot of figuring out, a lot of time, but that's the best lessons, though. The ones you never forget are the ones where you invested a large amount of time and then clearly failed.

(Sanjay at 00:32:09) Yeah. Yeah. You know, there's a big saying in the venture circuits that I'd take a CEO that has failed any day over a CEO that has only been successful because it's the failures that teach you the most. It's where the learning really comes from. And so, it's good advice.

(Joel Beasley at 00:32:26) You mentioned, you touched briefly on this concept of transparency within these AIs and algorithms. Are you seeing any tools that are out there that are helping make it more transparent about why the algorithms are coming to their decisions, how they get there?

(Sanjay at 00:32:44) Yeah. I think, look, I mean, AI fundamentally does have a little bit of a black box problem associated with that. It's just it's got good reasoning built in, but the reasoning is not obvious in a way that makes sense. One of the best ways we found to tackle that problem is actually to figure out the feature set, which is the variables that drive the end decisions.

(Sanjay at 00:33:02) So, you know, think about it as an example. We apply AI in a bunch of areas. One of the areas we're applying in banking is to balance sheets. So think about many banks will have loan portfolios that lend to small and medium or large business corporations. So you think about a loan that has a portfolio of a thousand small and medium businesses, and you have to understand the risk in that portfolio as a way to manage that business properly.

(Sanjay at 00:33:27) And, traditionally, the way that used to be done is you'd basically call people up and get them to send their balance sheets. You'd have a team of CPAs read the balance sheets and figure out what this company's got, so much risk, and then it's got this one, this other one here. Then you kind of aggregate it all together, and you come with the risk for the portfolio. I'm simplifying it, but you get the general point. Well, all of that takes a lot of time.

(Sanjay at 00:33:45) It takes a lot of, it's dependent on manual integration. You can't, like, react and say something just happened last night in the stock market, let's quickly reevaluate. Well, you can't get it done. It just takes time. And then, of course, it's expensive because you have to do it with all of the expertise that comes into play. Fast forward to today, we use AI. We apply artificial intelligence. We extract all of this at high speed.

(Sanjay at 00:34:06) We convert these PDFs into structured documents. We're able to extract meaning from these terms. We're converting them into numbers. We're applying formulas. And lo and behold, we can come with a spread or a risk score for that portfolio.

(Sanjay at 00:34:21) So what's the trick in that? The trick we've learned is if you just gave a risk score and say, well, this portfolio has got this risk, right? You get the question of, why would I, why is that right? Like, why is it not this? Why, how, why did it change? And the best answer you had was, well, trust me. Or, actually, the best answer you had was wait and see. Right? You can, you know, time will tell you, right? That's not a good enough answer and certainly doesn't work for regulators. So the trick we found was, can you breadcrumb the decisions through? And what I mean by that is, I'll give you a score.

(Sanjay at 00:34:53) You'll say, Sanjay, come on. Why is this score right? And I kind of go, you know, just click on it. And you click on it, and it drills down to the next cell. Now it shows you the 100 companies, and it shows you the risk number for each of them. The seventh one didn't make any sense.

(Sanjay at 00:35:05) Like, okay, click on it. You click on that, and it breaks down to something else. And you click on something else and keep, and you click and drill and you ultimately get to hear the two sentences we picked on page 17 in the footnote of this balance sheet. And here's a sentence. It just shows up on the screen. Right? You read that and kind of go, well, that's kind of interesting. Maybe it makes sense. What just happened?

(Sanjay at 00:35:28) What just happened is in the old world, I'd still be given a number. I'd still kind of go, well, you know, I don't know if that makes sense. Let me walk over to Mary. I go to Mary. Mary said, well, you know, that, so I walk over to John. And then John says, well, let me open this book and show it to you because, you know, that's what the balance sheet says. Right? So, really, this click and drill, this way of being able to introspect the number and be able to understand what feature, in this case, what is the footnote in this example that drove that decision, that feature understanding becomes really important. And so what we've done is we now implement AI with these breadcrumbs. You can follow the logic trail.

(Sanjay at 00:36:00) You can get to the point. And now you can go and say, well, why does Mary think that those two sentences imply something? Well, that's the same problem you had before. Like, because before you had Mary, who was making that call, she was a CPA, and you trusted her. Right?

(Sanjay at 00:36:12) So what we're trying to do is we're trying to emulate the thought process, the way you approach it to be more friendly, to be more nuanced in the experience of what you used to do so it kind of comes to life much better. But I think that's one of the ways we've found to solve for this problem of trust, and trust is important. But what's more important is adoption and usage. Right? The other thing is now there's new pieces of technology that actually allow you to look at something and then say, what are the features that are driving that decision?

(Sanjay at 00:36:43) And for regulators that are trying to understand whether or not to approve the use of something, it becomes very important because now they can look at the underlying components that drive a specific decision, and they can kind of follow that logic trail a little bit more. So we've made a lot of advancements both by, in the area of designing better for usage as well as in the area of this feature analysis that allows you to understand the underlying components of what drives a decision. And I think that's certainly improved the adoption and the usability of AI meaningfully for us.

(Joel Beasley at 00:37:17) Yeah. As you're describing that, I was like, that's, it's like we're taking human knowledge and then making it digital. Like, we're taking thought patterns or the way people work, like, these rule sets of how people think, and then we're digitizing them so we can deploy on that scale. And it's interesting because it's like almost the technological, or I think it was the industrial revolution. We just took the movements humans made and had robots make them.

(Joel Beasley at 00:37:42) And now we're taking, like, the ways people think and how we solve problems and then doing the same thing there. I'm just always curious. The way my mind always works is stretch that out, you know, two decades and what's the next thing we do. If we went from the physical to the mind, like, what's the next step other than, like, just neurolinking and going up into the computer?

(Sanjay at 00:38:05) I think it's a good question you ask, and it's one that, you know, all of us reflect on. I'll give you a perspective. I'll actually start with a quick story that I happen to like a lot. It's a story that someone else told me and it stuck in my mind, and I think I'm going to get the dates wrong. But let's say 30 years ago, roughly.

(Sanjay at 00:38:20) I'll explain why the date's important. Think about this. Five individuals used to get together every last Friday of the month at the donut shop around the corner for a cup of coffee and donuts, and they'd meet from eight to eleven. Guess what they were doing? They were figuring out the sales forecast for the next month.

(Sanjay at 00:38:36) These were five regional sales managers that founded upon themselves to get together and collectively think through trends and forecast better so that they could get their customers satisfied with the right products at the right point in time and not overstock and all this other stuff. The reason I said 35 or 30 years ago is because at the time, Microsoft Excel didn't exist. Right? So this is a verbal discussion. People get together with notepads and pen and all this other stuff, and it kind of worked well.

(Sanjay at 00:39:00) Well, let's look at the dynamics in that discussion. There are two people in the room that were just fantastic at arithmetic. They were good mathematicians. Someone would say, well, I think this thing's not going to work out. You know, this will come down by 2%, and this thing is going to be a little bit better or whatever.

(Sanjay at 00:39:14) They just figure out all the numbers in their head and come back with, alright, that means this is the number of units we need. Right? So they added a lot of value because if they weren't there, then you'd be spending a lot of time trying to work out the math. Two of the people in the room were actually good listeners.

(Sanjay at 00:39:29) They didn't really think as much about trend lines, but they heard everyone. They said, wow, that's a great idea. Well, this is going to change. This one I do buy into. And they'd be the best implementers. They take the advice. They tune the way they're sold. They talk about new areas, and boom. You'd see that reflecting the sales.

(Sanjay at 00:39:44) And then there was this one other person, right? The fifth person. And they just asked these kind of just these questions. They were just like these nagging questions. It's kind of like enough already. Like, what if this happened? Or what if that happens? Or what if this changed? And it's kind of like, you know, stop with the what ifs, will you?

(Sanjay at 00:40:00) Like, we've got to get work done. And so then Microsoft Excel comes out, obviously. That's why I said it's 30 years ago. And now in a second, in the flash of an eye, someone said something, you put it in, the entire spreadsheet recalculates. Right?

(Sanjay at 00:40:11) You know that, right? So now what happens to this group? Well, the two guys that were fantastic at arithmetic, their value is gone. Right? That whole thing isn't as important anymore. The two people that are out there that said, you know, once we make a decision, we're going to implement it better than anyone else. Oh my God. Like, now these become stellar because you have better decisions coming out, and then they're great executioners. They can make strategic decisions in the field based on a set of predictions.

(Sanjay at 00:40:37) But this one person that asked all these nagging questions, guess what? That person becomes the star of the group. Why? Because now you have this fantastic tool that'll recalculate numbers in a second. And what you need is the what if analysis.

(Sanjay at 00:40:50) What if this happened? Put it in. What if we did this other thing? Just put that in. What if we change this, that, and the other? Well, let's try that out, and then you can extrapolate. You can work through all these scenarios and come to the right conclusions. So the point I'm trying to make with the story is, you know, when new technology comes around, you know, we have a need. There's a need in the humans and the law to reconfigure themselves around it. And the arithmetic in this example doesn't make as much sense.

(Sanjay at 00:41:15) But the what if analysis, which is low on the list, by the way, it's a hassle, it was a nagging thing, right? Now that's, like, super important on the list of things to do. And this thing about taking some predictions and being able to act on it now becomes even more important.

(Sanjay at 00:41:29) And I think that's the view we need to take of artificial intelligence. AI is going to be great at prediction. It's going to make better and better and better predictions going forward. It's a prediction engine. That's what it does.

(Sanjay at 00:41:41) What's the probability something's going to happen tomorrow? What is the probability this other thing's going to affect? That's what it does, and it's going to get more accurate. There's no denying it. But once you have a prediction, you still have to make a call.

(Sanjay at 00:41:51) You still have to apply judgment. You still have to contextualize it. You still have to make a decision, and this is where I think humans play in the loop. And the more we think about it as that collective continuum, the more as we think about it as these are computers in a group, humans in the loop, right?

(Sanjay at 00:42:07) Like, you think about it as a collective ecosystem, and you say, well, this one kind of gets better. This one has to kind of compensate this way, and you reconfigure yourself around it. And I think that's the trick. I think too much of the narrative today is that, well, it's going to have this impact to jobs, and it's all going to be negative, and computers are going to run humankind, and da da da da. I don't believe in any of that.

(Sanjay at 00:42:23) I think there's an opportunity to use AI in a way that's really meaningful and is going to drive amazing amounts of productivity for us. Yes, that means we have to change the way we think. Yes, that means I can't go to this donut meeting and still do math and think that's what I'm going to do and be good at it.

(Sanjay at 00:42:38) I have to do "what if." I have to think by judgment. I have to apply that. And these are the sorts of things—I meet with CEOs and chief executives at many different companies, and it's one of the biggest discussions we have: What is the future of work? Right?

(Sanjay at 00:42:51) How do we reconfigure ourselves to take advantage of all of this technology in a way that we drive more value for our clients, more value for our shareholders? And that discussion is really crucial to it. So sorry, this is an area I'm very passionate about. There's many things that have been said about it.

(Sanjay at 00:43:06) So I know it sort of went on and off a little bit, but it's a really important topic.

(Joel Beasley at 00:43:10) No, I'm dazed out, just listening. Like, this is so good. Dazed is not the right word, but I'm just—what you're saying is true, and I like how you articulate this concept of the fear because I haven't heard that explanation before. That was good and that was new for me.

(Joel Beasley at 00:43:31) So thank you for that, because a lot of the fear-based conversation is the one that spreads really fast and that people like to have over and over and over. But I've often thought that what technology is doing is it's amplifying humanity. So, you know, it's amplifying the good parts and the bad parts. Right? It's amplifying everything.

(Joel Beasley at 00:43:56) We've never been more connected. I mean, we're bouncing light around the world right now talking in real time, which is magical and beautiful. But I think some of the things that people get upset about are how people use the tools and not necessarily the tools themselves. I think it's up to us to use them. We have to use the tools correctly, and then we'll be on a good track.

(Joel Beasley at 00:44:19) I don't think the tools are—well, I don't want to say this in case future AI hears me say this now. But I don't think the tools are going to—I think the tools will run us if we let them, but I think as long as we have competent, intellectual, intelligent, pragmatic people who understand how to balance things and discipline and ownership in these concepts. As long as we have really good, strong people in the world, then the technology will be the tool that works for us.

(Sanjay at 00:44:46) Well said.

(Joel Beasley at 00:44:47) I also like the idea of—or I'm curious on your thoughts. So do you think that we will get to a point where everything is automated or so many things are automated that we don't have to work as much, or work will change to be more creative endeavors or other human things that are more human-y, animal-like? Like singing. Singing is an animal thing too, right?

(Joel Beasley at 00:45:15) It's something we do. You were talking earlier about physics and the world and things that are physically constrained. I happen to—and I'm learning a lot about myself every day—but I happen to really like doing physical work. Like, I go to the gym every day. Right?

(Joel Beasley at 00:45:28) I like lifting the weights. I feel better than I—if I don't go, I feel horrible. When I do go, I feel great. And I like the concept that there'll be more time for that, but it's very unclear how we get there because you hear people talking to all these concepts of universal basic income and there's just so many ideas out there right now. And I don't even know if I'm making a whole lot of sense because my mind's kind of jumping around to everything, but hopefully it inspires a good response from you.

(Sanjay at 00:46:04) Well, you bring up a really important topic. Right? And we talked a little bit of the future of work and the distribution of labor. And clearly, it's going to evolve, and we're going to be doing less of the mundane and the repetitive tasks. And I think the trend line is going to be more towards judgment and making decisions in the back of those predictions. But I think the point you made about the role of the human as it relates to empathy and anything you talk about—human experience—but it's really empathy and understanding that is a really important part.

(Sanjay at 00:46:39) And I think there's a role to—and there's a need to take those creative elements, the whole element of empathy, the whole element of design. You talked about some technology earlier, but, you know, some of these things, you open a box, and the sheer process of unveiling and opening the box to take out, you know, AirPods or whatever, it just brings a sense of joy. Right? And understanding that and then being able to package that with the AI to make a composite product that has enough of the human and the empathy and the experience elements, but enough of the automation of the mundane and the routine tasks. I think that combination becomes very important. So I do think that as a society, we're going to have to find a way—and I think that's where the reskilling is really important.

(Sanjay at 00:47:27) But to find a way to actually reskill ourselves, to take on tasks at the higher level up, move one step up. You know, it's happened before. I mean, it used to be that data processing and training all this information used to be a big task. And then, obviously, with computers coming in, all that stuff kind of went into spreadsheets and went into Word documents, and that whole profession sort of got wiped out. But on the back of that, new professions started.

(Sanjay at 00:47:48) So, you know, who would have thought that graphic design was even a career pre-PCs? Like, that's a whole new—search engine optimization. Like, would that even—I won't even talk about it pre-Google. Like, what exactly is search engine optimization? Like, why does it even matter?

(Sanjay at 00:48:04) So the point is new careers, new ways of working will evolve. Some of them we know, some of them we don't even know, so we have to be very creative around that. And the role of empathy, the role of the humanness, the role of some of those human factors around experience become very important. And I do think that that trend line's obvious. You can actually already see it today, and it's just going to accelerate.

(Sanjay at 00:48:25) So I think there's a lot of new stuff coming at us. And so the question then becomes, what do you and I do? What do you and I do? And I think that—you know? And if you think about it, I think the answer there is you have to be very curious.

(Sanjay at 00:48:37) You have to be very open. You have to be able to look at opportunities and say, how do I reshape? How do I learn? How do I learn more to be able to reapply my judgment in an area that's new to us? And I think that's going to become a key skill set.

(Sanjay at 00:48:52) I don't think employers are going to hire for skills anymore. They're going to hire for aptitude. Not aptitude. Attitude. Do you have a curious mind?

(Sanjay at 00:49:00) Do you have the desire to learn? Like, that is number one on the list for when I try and hire people. Right? The things that are top of the list are, you know, are you curious? Number one, most important question.

(Sanjay at 00:49:12) Number two, are you collaborative? Do you have a high EQ? Not a high IQ. Do you have a high EQ? Can you work with a team of people?

(Sanjay at 00:49:19) Can you draw a set of diverse inputs? Can you bring a set of people together that won't normally be working together in a way that you can extract the best of their thinking and compose that into an end product? Then the skill sets of the future, you know, whether you're good at math, whether you are good at some other skill is less relevant. And I think that's the way we all need to start thinking about our own personas, our own learning, and our own development. And that's an important part of each one of our perspectives on life.

(Joel Beasley at 00:49:45) Nice. Yes, Sanjay. I want to be respectful of the 3:00 stop. So we've got about four minutes left. I'm just incredibly grateful when I get to speak to you because I love your insight.

(Sanjay at 00:49:59) You're very kind. Thank you for taking the time, and I'd love to carry on the conversation another time.

(Joel Beasley at 00:50:04) Yeah. You have an open invitation. So I'll check in with you next year. We can catch up and see what's going on, get more AI questions together, and it'll be fantastic because we only got through half the topics here. So I'm always excited.

(Joel Beasley at 00:50:19) So maybe next year, we hang out and have another chat.

(Sanjay at 00:50:23) Well, I love these conversations, and part of the reason I love them is because they're two-way. It makes me think and it makes me learn. It allows me to reflect on the things that come in. And to that note, I want to let your audience know that they can reach me or into me. I'm on LinkedIn.

(Sanjay at 00:50:36) It's easy if you get my first name, last name right, and that's not an easy task. But once you get it, it's easy to link into me, and I'd love to hear more from you. And if you disagree with something, please tell me, and I would learn more from that exercise. So with that, thank you again.

(Joel Beasley at 00:50:50) Yeah. Yeah. And we'll put everything in the show notes too. So we'll put your LinkedIn profile, a link to—is there any call to action for Genpact, by the way?

(Sanjay at 00:50:58) I think that Genpact is very curious and very much—I think there's two things I'd say. One is we're very curious about the future of work, and we have a point of view, and we're developing it iteratively with audiences. So one call to action is interact with us, pull us in, get us into discussion. We can help frame that together collectively. And I think the other part of it is the more kind of here and now, which is for people that can roll up their sleeves and actually get involved in kind of the dirty, detailed, grungy work of transforming core business processes.

(Sanjay at 00:51:31) And, you know, it doesn't mean that we have to work for you. You know, if you have ideas, if there are thoughts, you want to learn from our mistakes, if you want to just bounce an idea, you know, shoot us a note. Come to genpact.com. Happy to have that conversation any day.

(Joel Beasley at 00:51:44) It's like a kid in the candy shop what you guys get to do. I absolutely love it. Thank you so much.

(Sanjay at 00:51:49) No, thanks to you, Joel, and the team. Take care.

(Joel Beasley at 00:51:51) Talk soon.

(Sanjay at 00:51:52) Bye-bye.

(Joel Beasley at 00:51:55) Thank you so much for listening. And if you found this episode useful, please share it with a friend or a colleague that you think would get value from it. And if you have topics that you would like to hear discussed on the podcast, either add me on LinkedIn or send me an email: [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.