Navigating AI, Data Science, and the Future of Insights with Deb from Sigmoid Analytics

In this episode of The Greenbook Podcast, host Lenny Murphy sits down with Deb, Director of Data Science at Sigmoid Analytics and a 2025 Greenbook Future List Honoree. Deb shares insights into solving complex data challenges, the evolving role of Gen AI in analytics, and how companies are integrating AI-driven efficiencies. They discuss the future of data science, the shift towards strategic problem-solving, and the importance of upskilling talent in an AI-powered world. Tune in for a thought-provoking discussion on how organizations are embracing AI while navigating unknowns in the industry.

Key Takeaways:

• AI is revolutionizing data analytics, but strategic thinking and human expertise remain essential.
• Organizations must embrace unknowns and invest in future-proof AI strategies.
• The industry is shifting from traditional statistical models to AI-driven insights with LLMs and automation.
• The future of analytics will prioritize business impact over technical complexity.
• Developing strong problem-solving skills is critical for the next generation of data scientists.

Resources & Links:

• Learn more about Sigmoid Analytics

You can reach out to Debapriya Das on LinkedIn.

Many thanks to Debapriya Das for being our guest. Thanks also to our production team and our editor at Big Bad Audio.

Transcript

Lenny: Hello everybody, and welcome to another edition of the Greenbook Podcast. I’m your host, Lenny Murphy, and today I’m joined by Deb, the director of data science, Sigmoid Analytics, and also one of our 2025 Greenbook Future List honorees. Deb welcome.

Deb: Hey, Lenny, nice to be here. Thanks for having me.

Lenny: It’s great to have you. For the audience, both—it’s late for Deb, early for me, and we both have some type of, like, cold type thing, so we’ll get through, but forgive us if either one of our voices don’t sound… you know, [laugh] totally mellifluous. Anyway, Deb, tell us a little bit about yourself, kind of your role, and little bit of your background in the insight space.

Deb: Sure. So hi, my name is Deb. I work as the director of data science at Sigmoid. Specifically within Sigmoid, I look at the entire data science and the new [age 00:01:03] technologies practice. So, what that means is that I end up getting a lot of opportunities to work on some of the most complex and critical projects that our organization picks up. And I’m talking about unsolved problems where customers have, kind of, broken their heads for the last, let’s say, three years, four years. They could not find a product, any service company, or a consulting company to help them properly and give them the right results. So, yeah. I mean, that is a very interesting space to kind of pick up all the complex problems and try and solve them. What it also helps us do is we also get a sneak peek into what is next. What are the most challenging and complex things that the industry is facing? And if you find that there is some aspect of replicability that we can bring in, in the solution that we design, then that also gives us a first-mover advantage into building some of the cool concepts around new age accelerators. And what that also translates over a period of time is, if that tests success in terms of some thresholds, some benchmarks, some appreciation in the market, then we also kind of move them into some products under Sigmoid’s umbrella. So, this entire space of picking up complex projects, solving them, to building out the accelerators, to converting them into products is the space that I typically look at and support. Apart from that, I think overall, I have one-and-a-half decades experience in the industry. So yeah, I mean, I’ve seen quite a bit of things within the products and analytics services.

Lenny: Very cool. Thank you for that explanation. And yeah, it’s a wonderful business model, to actually address business needs, right, [laugh] and making money while you’re doing it. So, you mentioned getting kind of an early look at evolving challenges. Without giving away first-mover advantage, what are some examples? What are things that you’ve seen that really have been intractable issues for some of your clients that have prompted you to create some different products?

Deb: Okay, that’s a hard one to pick because there are so many. So, we generally pick up the top two or three in the organization. I would love to talk about all of those that we saw. But one of the recent ones that we’re solving right now is the application of Gen AI in a very regulated and complex space of med tech research. So, I mean, this company—just a bit of background—so this company has been trying to build some Gen AI application where their research team can accelerate their, let’s say, value propositions, the type of things that they take out in the market, how they kind of adhere to the FDA rules. But they tried to solve the problem with some of the biggies in the industry. They could not solve it for the last one, one-and-a-half years. Then Sigmoid came in. We are fairly through the journey right now where we are at a stage where we have proven the concept, the POC is successful, we are going for a scale up. And believe me, I mean, this is a very, very complicated space because new age LLMs, everyone loves them. They can interact, they can give you a lot of insights, but when it comes to med tech, these LLMs are not really, really trained to directly give you insights on top of med tech, let’s say, databases and documents and all those things. So, there is a lot of [craft 00:04:15] that we had to bring in order to solve it. So yeah, I mean, this is just one of the examples. I can just blabber and talk through more.

Lenny: Let’s actually dig in on that because I think the—so right before this call, I was reviewing some analysis done with deep research off of GRIT data, for instance. And you know, fairly high level, you know, there’s a lot of open ends and, you know, the stuff that we know that the AI is good with, but some statistical analyzes as well. Now again, not complex, but—and it’s a public, you know, just a general LLM—but you can see the writing on the wall, right? You can see, certainly with the appropriate amount of training and inputs, that those solutions can be incredibly effective in helping to solve complex analytical issues when given the right amount of data. I would imagine companies like Sigmoid are the folks that are helping to drive that specific business use case, that evolution of these platforms. So first, am I right that increasingly you guys are moving away from, kind of, traditional statistical solutions, you know, SPSS and R and, you know, all of those things, into LLMs as one of the components, and do you agree that we are moving fairly rapidly down a path where the interface that humans will use to engage with some of these will be more LLM-based, rather than kind of GUI, you know, numbers-based?

Deb: Really interesting question. So, the first things first. See, as a company, Sigmoid kind of solves a broad spectrum of problem statements. It’s not only about Gen AI; there are so many other problem statements in the entire analytics space that we solve. So, I mean, yes, we see that organizations are moving and are becoming more open to trying out these new age technologies, that’s true, but at least not in the immediate future, we see that classical analytics problem statements around forecasting, predictions, classifications, all those, are going to die down. Because those are also critical things that the organizations are trying to solve with more and more data that is coming in. Having said so, Gen AI is definitely adding another layer of efficiency to our entire organization as well. So, I mean, just diverting a bit into another aspect of my role here at Sigmoid, what we also try to do is what are the best possible ways to kind of upskill people, upskill data scientists at Sigmoid? So, what we are doing is we are trying to make everyone the best of the prompt engineers who are available in the market so that they can use those strategies, those techniques, to kind of accelerate in any project that they work on. And interestingly, I mean, these kind of ideas are resonating with some of our customers as well. One of the biggest companies in the, let’s say, baby food nutrition space in the world. Recently, I was having a word with the head of data science in their organization, and what he was trying to convey to me was that, “Hey, I mean, you guys are doing a lot of business with us. Gen AI is coming in. You guys are working on all the different business agents and bots that we want to build. Why don’t you find out opportunities—and that mutually benefits us as well—where we give you some money every month, every quarter, and you show me benefit in terms of how much of the traditional work that humans are doing in a less, let’s say, efficient way today, you can automate, so that we can put the same persons to work on something more meaningful within the organization? So, show me that value that you are driving, based on the money that we are giving you.” They want to definitely see a positive impact over there. So, yeah. I mean, interesting times, I would say. A lot is yet to be seen, but with the way Gen AI and this entire space is progressing, I think, yeah, I mean, we’ll be very soon seeing us a system where I don’t think a lot of knowledge, like a predefined knowledge, is going to carry much value because most of the things is going to be available at the just touch of a fingertip, or something like that. So, people are just going to access that knowledge and put that knowledge to application, and that is going to be an area that is probably going to have much more weightage than where we are currently.

Lenny: I agree, for what that’s worth. I’m not a data jockey about any stretch of imagination, but what I’ve even found recently for myself has been the ability to utilize those tools to very efficiently synthesize information to get to the answer to a business question, right, and an outcome, it’s a superpower at this point, right? The bandwidth multiplier effect alone is just insane, [laugh], you know? So, for myself as a—you know, at the highest level, I think we in the research industry, we are knowledge workers. That’s the very highest description. And you know, when knowledge is easily accessible from kind of a structural and process standpoint, which is what these tools do—which is a progression that we’ve been on for a long time, right? All the way—[laugh], you know, whether it’s Microsoft Office and Excel, or statistical packages, we’ve been progressing of this place, of getting knowledge easily accessible and increasing the utility of it. This is the next level. To your point, I believe that across the insights and analytics space that it is fundamentally going to be the skill sets required for success will be far more strategic in nature with the tactical layer, and that strategic component is: what is the business question; the second is, what data do we need to answer the business question; and then third, what’s it mean, right, giving good recommendations off of that. Is that what you’re seeing, kind of a similar progression overall?

Deb: Yeah. I mean, pretty much bang-on that one. That is what exactly all the organizations are striving to do. In fact, if I just add a layer on top of that, so organizations, and let’s say the head of data, the CXOs, are becoming smarter by the day when it comes to how they want to adapt some of these new age technologies. I mean, rarely do I see today, which is, like, a very, very long-term plan and a strategy, a data strategy, which is not tracked very closely at a monthly or every two months level. And that is where people are, I guess, becoming smarter. When I say that, they now understand that we are not doing magic. We are actually talking about implementing something. Yes, there are going to be unknowns. Things may fail, but what is more important is for people to kind of continuously track, do those course corrections have a shift in direction, if needed at all? But yeah, I think tracking has become much, much more, let’s say, progressive over the last years, I would say.

Lenny: All right, so you, in your role, right, you are leading the data analytics strategy around Sigmoid. So, as we enter in this brave new world, faster than I think any of us really anticipated it progressing, what are you focusing on as kind of the secret sauce, if you will, for Sigmoid, and what do you think that means for the industry? I mean, obviously you don’t want to give away, you know, the competitors, give them [laugh] the keys of the kingdom, but kind of step back as a visionary and say, here’s what we know this is going to be true, and this is what we’re focusing on within our teams, within how we collaborate. You know, what’s that look like?

Deb: So—I mean, see, I mean, like you said, like, there are secret sources, right? Every organization has, and there are directions that we are chasing, for sure. But I think there is one common thing, I think every organization is striving to be at the best possible level that they can, specifically whenever they are in the consulting and services space. I mean, this particular space, often, the success in this particular space is often directly proportional to the number of good people, good knowledge workers insight workers that are available in the organization. So, I mean, as a company, Sigmoid is trying to solve this problem where, I mean, we know that if we go by the law of statistics, there are only a few best always in you—build in any distribution, there are only going to be a few best. How do we as an organization be at a stage where we can perhaps pick up some good people and convert them into the best possible in their particular line of work, or in their specific area where they want to kind of excel and grow? So, that is not a very easy problem to solve when you are talking about training people, but that is one area that we are chasing crazy as a company, where we feel that the scale of our business should not be directly proportional to the best number of good people we find in the market, but it should be—I mean, we should be in a stage where we can get good people, train them to become the best in a specific aspect. So, we are doing a lot of brainstorming, ideation, putting out frameworks to test in order to test some success over there. It’s an ongoing journey; we have tested some success, some frameworks which are holding good. But yes, a lot is yet to be seen whether we are able to bridge the gap or not. Otherwise, it becomes highly difficult for any company like ours to kind of school. We are now 1200-plus in the world, globally. We are talking about 2x to 3x growth over the next two to three years. How do we find 2x to 3x additional the best possible people in the market? That’s a real problem to solve, I guess.

Lenny: So, do you think the identification of talent in this space is going to get easier or harder? And let me build on that for a minute because the cynic in me—and actually even as a parent, so I have kept my children away from AI. It’s like, nope, you got to build these traditional thinking and solving skills like I did, you know [laugh]? You got to do the work, while at the same time, I am increasingly offloading some of the process-oriented tasks that I’ve done into AI so I could do more value-added. And I see a tension there with—I could see a point where the real skill sets that we have to teach the candidates of the future are basically critical thinking, while also needing the foundation to be able to validate outputs and truth. I mean, I’m sure you’re seeing it. No matter how good I think these tools are, I still got to double-check, there’s still hallucination, and errors, and all those things that occur. But yet, our education system, even the STEM programs, I don’t see them focusing on that thinking component; I see them focusing on the process component. And is there going to be—is that going to create challenges for us as an industry, or for you, for Sigmoid, as we’re recruiting more people? Or am I totally off base? I may be totally off base. What’s your take?

Deb: The problem is almost similar, irrespective of whether we are Sigmoid or whether we are working with someone else. We are talking about a common problem in the industry, that the number of best people is always going to be limited, but at the end of the day, what matters are the problem-solving skills. And that is what we are trying to bake into our system as well. When I say that we are trying to follow frameworks, we are trying to structure the entire org in such a way that not everyone needs to learn how to solve all types of problems. The way we should probably create this is that a person becomes the best in solving one type of problem first, then he moves to solve the next type of problem. And how can we accelerate someone’s journey in the process of solving that first type of problem? Once a person knows how to solve, let’s say, one type of problem, the second type of problem, they automatically start to get into that mindset. They get a lot of those, let’s say, hacks how to scale and how to solve different types of problems, let’s say. I mean, that is one critical thing. I would say that getting the best people is always going to be a challenge. The focus should be always on how can we get some of the good people, identify where they can add value in the fastest possible time, which axis of problem-solving they can kind of learn faster than anything else, and then push them towards to the right set of techniques and frameworks so that they become experts in that first before they try to crack something else. As an organization, this becomes more challenging, I agree, because at the top level, we need to kind of aggregate all of these things. At the top, I can’t just pick up a project and say, I’m going to pick five people into that, or ten people into that, whatever that is, right? At the end of the day, client is going to pay x, and we have to kind of make things happen within that x only. We don’t have a luxury to bring in a lot of people just because the first person can solve one type of problem, the second person can solve another type of problem. So, at a high level, yes, this is a challenge at a strategic level to solve, but I think we will kind of be at a much better place in the future if we kind of groom people in the best possible way because these are the people who are going to kind of solve these strategic problems for us over the next few years, once they stay in the system, once they grow.

Lenny: Now, does that—and I think that makes perfect sense—does it also reflect your own experience? So again, as a Greenbook Future List, you’re relatively younger in your career compared to an old fart like me, and you’re looking at, obviously, the next generation. Has your own experience informed that thinking of mentorship, and you know, focusing on the right skills for the right job and the right person? Tell us a little bit about that.

Deb: Yes, I mean, with every generation, I mean, when I look at some of my previous generations and when I look at some of my younger generations, I mean, I learn a lot about how they kind of look at a particular problem, right? We get to see different lenses with which they see a particular problem. Having said so, I mean the younger people, I mean over the last five years—if you say five to ten years, if I put it like that—the younger folks are now getting trained in a certain way. You have AI crash courses nowadays, I mean, which claims to teach you AI in a week or in a month’s time. I mean, all of that is good. There is a lot of, let’s say, readily available information through which someone can get to, let’s say, a particular level which is not level zero, definitely more than that, in the fastest possible amount of time. But is that enough? That is a question we need to solve. And often, I mean, there are downsides to this as well. When people are just thinking that, yes, I can pick up a skill set and I can become the expert in that within three months or six months, or however these programs kind of advertise, I mean, that kind of kills the overall growth mindset in them as well because sometimes they just do a course and feel I have achieved something in life, and I am the best data scientist available in the industry. So yes, I mean, we are talking about an intersection of human psychology, to a certain extent. We are talking about how fast people can learn, and new ways of how younger folks looks at the same problem. It is becoming a hard playground to play in, but I guess this is a solvable problem because, at the end of the day, with the right mentorship, feedback, people learn—the best people learn. The not so good people kind of live it in between. And we can separate out whom we want to move ahead with, to put more efforts to [unintelligible 00:20:26].

Lenny: All right, so I want to kind of look towards the future now. So, we’ve been talking about this transformative state that we are in, with kind of a short-term view, but as you look towards maybe the five-year horizon for the industry as a whole, what do you see it looking like? What is the shape and contours of the industry driven by all these changes that we’re talking about, the new skill sets, and generation, and AI, and data ubiquity, and all of those things. What’s your take?

Deb: So, I think it will be hard to predict the next five years. I mean, we are trying to predict what is going to work in the next one to two years, even. It is that fast nowadays. But yeah, I mean, at a strategic level, if we look at it, generally, most of the senior folks create strategy at a much higher level, and what I feel is, that organizations are already understanding that they have to invest in unknowns so that resistance to invest in unknowns is decreasing. They’re thinking of new challenges, and they’re okay to fail, as long as they are failing fast, and they are not just learning it after they are kind of burnt a lot of money into that, or put in a lot of efforts into that. I think this is one trend that organizations are going to pick up much further over the next five years, that they will invest and invest more in the unknowns, with an eye on the possible benefit, the reward that they’re going to reap if that is successful.

Lenny: I like that because there’s the known unknowns and the unknown unknowns, right, with those contingencies. And I think that we are in the era of what feels like fairly constant black swans. I’m not sure if everything really qualifies as a black swan, but unprecedented change, unprecedented pace of change in things that often seem to go—you know, people will say, “Oh, I didn’t see that coming.” And whether we’re talking—at all levels, right? You know, geopolitical, and technological intervention, and you know, there’s just [laugh] so many things that are just really interesting times, and that contingency planning, that thinking about those changes, is really interesting. Because, personally, disruption means opportunity. It’s still going to be disruptive. It still may be, you know, not fun going through that process, but on the other side, there can be really interesting opportunities. Are you seeing clients actually embracing that and trying to, with that ethos, look, here’s our business today, here’s this whole laundry list of contingencies, of unknowns that could happen, and actually taking that to the next level on, and here’s how this could open up new opportunities for us, and having an allocation set up. You know, here’s the probability of these things happening, and there’s a plan that’s almost, I think of that type planning as military almost, right? Are you seeing people really embracing that, or is it still a little bit of a foreign concept of but wait, we just want to know our, you know, our net promoter score and, you know, [laugh] our brand tracker. What? What do you see happening?

Deb: I mean, like I said, like, people are kind of becoming more open to trying out new ideas. I feel that they are just going to do that more if they have the right level of trust. I mean, as people from the consulting and services industry, I mean, it kind of also comes down to us, the kind of trust that we can provide them with. But yes, I mean, anyone who would love to see some success first before they actually invest everything into that, and that is going to hold true everywhere. Like, someone has to earn trust. In order to earn trust, yes, we have to show them some benefits, some, let’s say, good results coming out before they actually take a decision to invest. One specific way of working with clients that is opening up nowadays is, they are kind of trusting a lot these external companies to, kind of, help them navigate the unknowns within their organization. So, they are becoming more open in that sense. They know that, yes, there are unknowns to us, and in that sense, we should get someone that we can trust to come and help us in order to, kind of, solve those unknowns. But they are also becoming more and more accommodating, if I have to put it like that, to understand that even if I bring in an expert—let’s say you asked me a question that what is going to happen in the industry after five years—like, as an expert, I may not be having the right answer. So, they are becoming more accommodating when it comes to respect that, yes, the expert can have unknowns as well, and jointly, the expert and the organization should work together. I mean, we are seeing customers trusting us to kind of build data labs, innovation centers, on behalf of them, within their organization because they want us to be a partner in their journey, in their roadmap, rather than just being that, let’s say, wonderful company, to kind of just solve a problem, and then move ahead, and pick up something else and then solve it. They want us to—I mean, they want these consulting companies, all the experts, to come in and help them build expertise within the organization. So, that’s a new thing that I see.

Lenny: So, as we kind of wrap up, let’s talk about you as an individual. So, outside of work, what passions or hobbies do you have? How do you spend your time other than trying to solve big, chunky data analytics problems?

Deb: So, I mean, somehow I feel this chunky data science problems are my go-to passion. It has been my go-to passion for the last ten years, but there are definitely more things that have added up in my list. Like, I’m in a state right now where I like to travel. I like to travel. I like to live the life of a, let’s say, digital nomad. But at the same time, I like to solve the same problem, not sitting down in one location, but maybe travel across different locations and keep solving that problem. So, yeah, I mean, I like traveling. I do a lot of road trips. I like to drive. Like, adventures is driving, if I were to put it like that. And I have also picked up a knack of water sports nowadays. So, my wife, kind of, pushed me into this almost a year, year-and-a-half back. Prior to that, I never, kind of, participated in any of water activities, any water sport, but since the time I did the first one, I mean, it has been, kind of, like—it gives me an adrenaline rush, and that I really enjoyed. I keep trying, keep trying new ones, wherever I see it. So, yeah, I mean, these are some of the things that are going on in my life.

Lenny: Okay. Very cool. I like road trips as well. It’s so much more interesting just to take the back road, and drive, and see what’s going on in the world than just flying over it. So, although I think it’s probably a little more of an adventure in India than is in the US, simply because of the driving, the other drivers on the road, from my own experience [laugh]. So, that is a—if you’re a road trip adventurer in India, your skill set is far higher than most. So, I give you props for that, my friend.

Deb: That’s true. I can’t agree more. Yeah.

Lenny: Yeah. So what’s, uh, what’s next for you? Anything that you’re just really excited about for you, just personally, thinking about your career, your family, whatever the case may be?

Deb: Me as a person, I mean, though I kind of consult everyone on going through different changes, but me as a person, I’m really averse to changes. I like how things are right now. Yes, I mean, in my mind, in my life, I’m chasing some goals; there are some bucket list that I’m trying to chase, but the beauty of having a bucket list is that, at least with me, that bucket list never remains the same. It gets updated every year. So yeah, I mean, I’m just happy with the surprises that are coming along my way. I’m just happy with the way I’m navigating through those surprises, goods and bads. I mean, let’s see where life takes me.

Lenny: Good. Well, there is nothing wrong with being content with your status. So, that’s fantastic. Many of us strive for just being okay with how things are, so that’s great. Congratulations again on being a Future List Honoree. I am sure that we will see and hear more great things from you as things progress. Where can people find you?

Deb: The best way to connect with me is the only social platform, I mean, is LinkedIn. So, that’s the best place to kind of catch hold of me or reach out to me. Apart from that, I guess, the social medias do excite me a lot. My contact and my email id, everything is there on my LinkedIn. So, I mean, yeah, I mean anytime, I’m just a click away in that sense. Just drop me a note, and then yes, we can start chatting on anything interesting that catches us both.

Lenny: Great. Thank you so much for joining us today. I think that was a great conversation. Hopefully everybody got a lot out of it. I know I certainly did. You gave me much to think about. I want to give a shout out to our producer, Brigette—she keeps all the balls in the air—to our audio editor, Big Bad Audio, and to our sponsors, and most of all, to your listeners because without you, Deb, and I wouldn’t have—we probably would have hooked up at some point, but this was a, you know—it forced us to have a great conversation, and I think that’s useful for everybody. But that’s it for this edition of the Greenbook Podcast. We’ll be back with another one real soon. Thank you all. Take care. Bye-Bye.