Evolving the Enterprise

Data and Analytics in Retail: Navigating Walmart's Data Landscape With Amit Shivpuja, Director of Data Enablement at Walmart

• SnapLogic • Season 3 • Episode 3

Share episode

In this episode, we dive deep into the world of data and analytics in the retail industry with Amit Shivpuja, Director of Data Enablement at Walmart. Amit shares valuable insights and experiences from his role, shedding light on how Walmart leverages data to enhance decision-making, drive efficiency, and foster a data-driven culture.

Throughout the conversation, Amit discusses Walmart's data governance framework, emphasizing the critical role of data quality, consistency, and collaboration across functions. He also explores the evolving landscape of data analytics and the impact of automation, including generative AI, on streamlining data-related processes.

Listeners will gain a comprehensive understanding of how a retail giant like Walmart manages its vast data resources, fosters cross-functional collaboration, and harnesses the potential of data to drive business success. Whether you're a data enthusiast, business leader, or simply curious about the inner workings of a retail data powerhouse, this episode offers valuable insights into the world of data and analytics at Walmart.

Sponsor
The Evolving the Enterprise podcast is brought to you by SnapLogic, the world’s first generative integration platform for faster, easier digital transformation. Whether you are automating business processes, democratizing data, or delivering digital products and services, SnapLogic enables you to simplify your technology stack and take your enterprise further. Join the Generative Integration movement at snaplogic.com.

Additional Resources

Data and Analytics in Retail: Navigating Walmart's Data Landscape

Dayle Hall
Hi. Welcome to the podcast. I'm Dayle Hall, CMO of SnapLogic. This podcast, if you've listened to any of the previous ones, we're here to give organizations some insights, some best practices on how to think about integration, automation, and how to transform the enterprise.

Our guest today is Amit Shivpuja. He is a BI and analytics leader, 20 years of experience, very passionate about driving data-driven growth across enterprises with an interesting holistic approach around people, technology, data, algorithms, processes and governance, the whole gamut of running data within a large enterprise. He is currently the director of data governance and strategy at merchandising at Walmart, that small company that probably many people heard of.

I know it's Amit, but I'm going to refer to you as Shiv because you gave me that. So Shiv, welcome to the podcast. What I like to do originally- initially, is to just give me a little bit about yourself, how did you end up within this role, you had a passion for data in the past? Just give me the two-minute cliffnote, Shiv.

Amit Shivpuja
Let me answer your second question first, and I'll come to the question on how I ended up here. The intro to data was purely fate. I just completed my undergrad in computer science, was looking for a job and I got into Robert Bosch, the automotive supplier, and was picked by the car navigation systems department. And my team, we would buy third-party datasets and then customize it for the GPS maps and addresses and all that went into car systems at the time. So that was my exposure to data. So I didn't really plan that software engineering, I have to do data. It just happened as fate would have it.

But coming back to your first question, which was, what was my journey like and what got me here is, in the past 20 years, I've had a very unique journey with data analytics being the central theme. I built pipelines and data flows as part of Bosch, then did my MBA because I realized I liked management a little more than technical. And then after that, I have product managed, business developed, marketed, sold, consulted, and then been brought into building grow data and analytics teams, which is what I've been doing for the past couple of years. So I've taken a very different route to be in the data and analytics space. It's also a more business route than the more- the pure technical route that people take. And it's been with- like when my- after MBA was with Alteryx, which is a very popular BI solution, then I sold Oracle for a little bit, OBIA, OBIEE and the like. Then I joined Deloitte and built the location analytics practice, and then did virtual reality for a little bit.

So I've had a very diverse career, thankfully, 20-odd domains, different sizes of organizations, fintech most recently. And now, Walmart has been thinking about data very strategically. And in order to enable that strategy, they’re focusing quite a bit on data governance and strategy, whether that's stewardship, whether that's data as a product, whether it's figuring out those strategic capabilities that they need to succeed in the future. And each function has this kind of capability. So the merchandising org, which is the one responsible for procuring all of the material that Walmart sells to consumers, was building this team. And this is in my journey to becoming a chief data officer at some stage, the focus on data governance and strategy specifically appeal to me as a role. So here I am at Walmart.

Dayle Hall
You do have a different background. A lot of people that I've talked to on these podcasts, they've come through engineering or maybe like designs or a very different route, but I like the experience you have because you've been within different parts of the organizations, the line of business, that gives you a unique perspective. So is that something that you feel- how do you feel that perspective helps you in your role around data governance? Because, again, it's an interesting way to approach the role. What special skills do you bring to this type of role?

Amit Shivpuja
I wouldn't say special skills greatly, but I would say that it gives you the end user perspective.

Dayle Hall
I like that. I'm an engineer, generally. So I like the fact [from that] perspective.

Amit Shivpuja
If you think about it, I’ve always thought of analytics or data science and all the test tools. Think about it like a toolkit, a hammer or a spanner. I'm oversimplifying, but you get the idea. But you need to apply it to get value. And that direction or the focus to apply it is business context or that end user perspective. So I think that sensitivity, that ability to more easily put myself in my stakeholder shoes, despite having been in so many domains, has helped me out so far.

Dayle Hall
No, I can- again, I can imagine. We've talked on this podcast before. I've had conversations with analysts for the last 10 years around how important the lines of business are going to be with IT purchases, with applications, with data and so on. And we also talk a lot about things like shadow IT, where either businesses are trying to live- they want to move fast, and sometimes they feel that IT is maybe not supporting them as quickly as they can. So I like- that must give you a unique perspective when you're approached with how to support the lines of business.

Amit Shivpuja
One of the top pieces that I had an opportunity to put out, I call it the sandwich for making data decisions. And the reason that I call it a sandwich is the two slices of bread, one on top is the business domain and the one below is security and data protection, right? But the reason I call it- I named those two as the bread is because they define the outline of the sandwich. You can use data science for healthcare, but then healthcare defines the outline of the sandwich, right? So that's what business really does. It gives it that focus thing. Otherwise, it becomes theoretical, right? I can run the best AI algorithm in the world, but if I don't have a business outcome that it's benefiting, it's a theoretical experiment.

Dayle Hall
Yeah, for sure. So talk to me a little bit about data governance then at Walmart specifically, or how it plays a role across the organization. Because I know you're in a specific part, but how does Walmart think about data governance? Are there multiple groups? Do you sit on councils? Like how is that run in such a large organization?

Amit Shivpuja
So let me just set a little bit of the context, and then I'll dive into structure. What surprised me is how maturely Walmart has been thinking about data holistically. I'm saying this because I talked to peers, I've seen the challenges we have, you get to meet people in conferences and all that. And you realize that even huge organizations who have invested a large amount of resources are still in very early stages thinking about data as an asset, or something of importance, right? I like to keep saying that it's often a stepchild. They think of the products, they think of everything else, but data is the stepchild. I think Walmart has a hybrid vision around data. There are certain things that are centralized. There are certain things that are democratized, so to speak. So you have core capabilities like compliance, the data management ecosystem, data engineering teams, that are very centralized structures. So there's one huge pool of data engineering resources, so to speak.

Across all of Walmart, there is one central data management and compliance. We call it data citizenship organization that sets it for Walmart as a whole. But the- so you here are defining like the basic infrastructure, the foundation, which is the data pipelines and all of that, and then the compliance and all of those guidelines. But then when it comes to tailoring it or customizing it to the different business groups’ needs where you need some domain knowledge and you need to tailor the compliance or you need to enable those data assets, that's where it gets decentralized. So we have- so like my department, each function has its own data enablement functions. And data enablement is basically three things, stewardship, data as a product, and strategy. And we are a business team. We are not an engineering team. We report to business.

Dayle Hall
Okay, so you report to the lines of business, so you’re there to enable them to be more successful. So if you're- on these types of podcasts, I always try and think about, okay, so if someone is out there listening to this, they likely are smaller organizations that [inaudible]. Do you find that- what are the benefits of having that specific data enablement function of some core functions? And would it- is this as much as your knowledge from the past- because I know you haven't been there too long at Walmart, but are there challenges with that? Were there challenges put in place? Because, again, I think every organization tries to create their own model. But I think what we're talking about is, I like core functions, data enablement sits on top with [inaudible] business experience. So what works really well with that and what are some of the challenges you face?

Amit Shivpuja
So the advantage here, especially for a data enablement or a data governance structure, is the closeness to the stakeholders. You're very sensitive, you're embedded or you're in the ecosystem of the business that's happening. The other thing that it also does is that you also then have the ability to specialize. Because it takes time to build that knowledge, right? You don’t- not everybody just knows. Even if I know supply chain for many years, it doesn't mean I know how a Walmart supply chain works for many years, right? So picking up those nuances and all of that and being closer to the stakeholders is also critical. The other thing that it does or enables is it allows best practices to be really customized the way that they should be customized. It's not a cookie-cutter, everybody follows this, take it or leave it, hierarchical thing.

Dayle Hall
I was just thinking, do you get- because you're in a specific function within Walmart and you'd sit apart from some of the central functions, does that- the people within your, call it, line of business or within your organization, when they're looking to do something, they approach you first about that. So you're a trusted partner to the business and you can get the right resources, that's an interesting model.

Amit Shivpuja
Correct. Elaborating on what you just said is, we then go look up, pick up what are the standard company policies, compliances, all of that stuff. We then look at our engineering resources, pick them up, and then we combine that to deliver the data that's needed for the same. Now, I'll be honest, right? The disadvantage of this kind of a distributed structure, just generally, is you have to invest a lot more time around things like documentation and knowledge sharing. Because the more specialized resources get, the harder it is to move them around. Let's say someone says, hey, I've been doing this for five years, I want to go do something else. If you don't have that disciplined approach of documenting, stewarding and capturing things, then you don't want to let that resource go because all the business knowledge is with those set of resources.

Dayle Hall
Yeah. I think documentation in most organizations is always a challenge for your type of organization. But what I find interesting about what you said, and we just talked- I just mentioned it, we talk about or you hear about shadow IT. And what I like about the approach that you and Walmart have is, to some extent, I don't want to say it normalizes shadow IT, that's not right. But you sit between some of those core functions and the line of business. So it's almost like you're embracing the opportunity to support the business, but also be the central point between some of those core functions.

Amit Shivpuja
So just to add something, actually, if we do our job, you shouldn't need shadow IT.

Dayle Hall
Yeah. I think you're right, absolutely right.

Amit Shivpuja
Yeah. Because shadow IT comes up because business doesn't feel the needs are being met.

Dayle Hall
And they don't feel people in your type of role understand the business.

Amit Shivpuja
Exactly. So if we do that specialization, if we do that collaboration, if we do that proximity the right way, what we are doing is, we are putting our- we are the business representatives to engineering, and we are the engineering represent- I like to say we are a bridge. We take business stuff, translate it into data and analytics, and we are data and analytics and we translate it back into business stuff. So we are the translation bridge is what role we play. And I think that applies to anybody in the data and analytics space, right? To be successful, that's what you want to do. But there's another thing that is very possible here, which is, we can think about data strategically. What I mean by that is, we can think about data long term. So let's say a team comes and asks for a dataset, it's not in place, we need to do some data engineering work. If we were very monolith, then we only see it from their perspective. But because we are distributed and we are horizontal in merchandising, for example, in our case, we can actually see other teams that would benefit from the same data and build the data once with that vision in mind even if it is a longer timeline. Because in the future, it gives us benefits. And this is I think one of the core things about why data governance is necessary, is to think about data strategically and not tactically of, hey, I want to build a dashboard, or I want to do this analysis, or I need this column, but, hey, what could they need and how could they need it so that we make it as reusable as possible?

Dayle Hall
That's interesting. I like the concept of- and I think data- you hear these terms now, right? Data is the new oil. And we talk about a lot of organizations are now trying to leverage data to be an asset. When you hear about data as an asset, what does that mean and how do you think about that permeating across Walmart?

Amit Shivpuja
There are a couple of perspectives to this, right? If data is an asset, one, it takes a village. What I mean by that is everybody has a role to play for it to be an asset. Right from the people, the engineering teams who build the products that generate the data, the product teams want to measure something, the business teams want to see value, to the data teams like ours that enable that, all of us have a role to play, whether it's data quality, whether it's data engineering, whether it's data modeling an architecture, whether it's dashboarding and analytics, so all of that, right? So that's one of the things that stands out to me when you say data as an asset. Two is, you'd have to know the value of your data.

Now, I don't necessarily mean it has to have  a dollar amount, though, that's one of the hot topics right now amongst data leaders, is how do you get to a dollar amount, but the thing that you need to make a conscious value-based decision on whether you invest in that data or not. Should I even collect that information, what difference does it make, etc., right? This also factors into things like the storage costs, processing costs, right? Data maintenance costs, all of that as well, right? So there is a financial implication to that. The third thing when someone says data as an asset, for me, is that it is due a certain amount of respect. You have to treat it like any other asset of the organization, you have to protect it, you have to make sure the right people get access to it, you have to make sure that it's maintained and the like. So that's where it kind of butts against my analogy that I gave previously of it being a stepchild because not everybody upfront thinks of data, they think about it later.

Dayle Hall
Yeah, for sure. And I think there's definitely been a shift over the last decade around- I think people understand that having the data can give a lot of value and insights to the business. But I still feel like in a lot of organizations, they're still trying to figure out, you mentioned it yourself, what is the value, how do they really put that into- is it monetary, or is it how it helps accelerate product development, or open a new market? So how do you think- if I asked you like, how would you- how do you class data, how do you class the value of data, what is it that you focus on specifically at Walmart?

Amit Shivpuja
I'll give you a partial answer because this is something that, like I said, we are all trying to figure it out, right? So if you take it top down, I think you can layer it in terms of, hey, what's the business value of the initiative that you're trying to enable using data? What's the product value or the product ways of getting benefits from this data as well as what are the actual operational or tactical benefits you get to time savings, reduced risk, etc.? So I think that's one tool of value. The other tool of it is a little more technical, and that's the more bottom up one. How much was it going to cost to store this, or how efficient is it to process this or get this data, those kinds of operational costs that come up, which is more bottom up, right?

When you want to enable this ecosystem, I want so much performance, I want it synced every hour, I want it real time, all of those things. So I think my theory is, the sweet spot is somewhere in between that, where you come up with the thing. Now, another perspective that I've heard of, which I'm thinking about right now, is you actually have a business outcome that you can connect it to, which is part of that top down thing I was talking about, which is, hey, what's my cost of acquisition, or LTV, or savings in terms of our increased subscription revenue, or whatever that is. But again, that's only the top down piece of it. There's a component of the bottom up, which also needs to be factored into it.

Dayle Hall
I think you're right on both of those areas. I think if you can show the value of data on the business side, I think it has a big impact. But the operational piece, do you think we'll ever get to the point where people are looking at what they have because we're not going to be generating less data over the next day? Do you think people are going to get to the point where they start to make decisions around the operational cost of capturing it, storing it, using it, it’s just become prohibitive? And at what point does an enterprise make a decision?

Amit Shivpuja
So we have to nuance that, the square, there's the answer to this a little bit. I think, honestly, every company should capture as much data as they can. I say especially for interfaces of public-facing stuff, everything a customer touches should be captured. But when it comes to the actual usage of it, the actual leveraging and enablement of that datasets, that's where I think it has to be driven by business need. Now, if you look at even some of the compliance things like GDPR, GDPR especially, GDPR challenges companies saying that, hey, why did you collect this data? So thinking both from, hey, what business value can this data give versus also thinking, hey, what's it going to take for me to get this data, is it part and parcel of what I'm doing, or is it going to take additional effort to get hold of this, right?

So again, it comes down to the top down thing. But I think where that sweet spot comes in is to have those conversations happening at a constant pace. You can't have the things in silos. So one of the things that I believe in is to try and permeate business knowledge as much as possible to the engineering folks, as well as educate business people as much as possible on the technical challenges or the technical ecosystem. Because business sometimes thinks, oh, give us the data, what's the big deal? It's there. But they don't realize the effort that it takes to get there, the pipelines, the coding, the quality steps, all of that. So I think that constant education and being that this is also part of that bridge I spoke about, we have to be that bridge on enabling the flow of information both ways.

Dayle Hall
Have you engaged in some of those conversations with your business? How does that go? Because again, look, I'm a marketing guy, right? Yeah, I want customer data. I want web analytics, product telemetry. I want a ton of data. But do you find that they're willing to sit down and listen to the challenges because they know it will give them the right outcome?

Amit Shivpuja
So that, again, is an interesting conversation because it depends on the culture of the organization. Luckily, in Walmart, there is a very data-driven culture. So what that means is that not only are leaders talking about it, but also even junior people realize that, hey, this is making an impact to that team over there. Like my data engineering teams know the business value of many of the projects where we enable datasets. It's transparent. We challenge our business stakeholders to say, hey, give us the business value before we start working on this.

But to answer that question, when we have those conversations, we have very transparent and open conversations on, hey, this is the technical steps we need to take in order to do that, which is why it takes so long to do it, or we don't have the technology today and we have to do this, or the other way around, which is, hey, this is the business steps that they have to go through and this data will help automate that. Yeah, we have that. It does take a little bit of education. It does take a little bit of negotiation. It does take a little bit of sharing, knowledge sharing. But I think what's happening is, like you rightly said a few minutes ago, is companies are changing their perspective from thinking about decisions purely from a product perspective, etc., to also saying, hey, I need to think about the data.

Dayle Hall
Yeah. No, for sure. It's interesting, what you said is that Walmart is a data-driven culture. And one of the things that we talk about at SnapLogic, and our CTO is studying for his doctor or masters, I can't remember. He's got so many credentials, it's [unfunny]. But one of the things that he talks about is, you can have new tools, you can [put something] like SnapLogic in there, you can change processes. A culture within an organization is also a key third element of leveraging data for those best results and so on. So give me an example like, how does that work in Walmart? What is an example of it being a data-driven culture? Because it must come from the top.

Amit Shivpuja
Yeah. So beyond the usual thing where leaders talk about data, you actually see them using data. That's one of the low-hanging fruits. It's when the C-suite, the VPs, the SVPs, the directors, all of them are like, hey, this dashboard tells me this and therefore, this is what's happening, and therefore, that's what we're going to do, right? But another key part of that is where they hold us to those data needs and standards and quality, the amount of push and investment they do in order to get the data that they need to make those decisions. So that commitment is the top down piece of it. The bottom up piece of this is where we figure out the best solutions, the best way to do it, the most performant systems, the most how do we minimize data quality from source, steps like that so that we meet that confluence point of delivering the data that's needed. So you don't want a scenario where there's no trust, where data is delivered, but nobody is using it. I think if you don't have trust, then that's the end of the whole data journey.

Dayle Hall
Yeah. I like what you said, though. I like what you said like, leaders can talk the talk. But if they then use the data, and it helps to make decisions, and they're very transparent with that, I can imagine that permeates down through the organization. They feel like, yeah, we are a data-driven culture.

Amit Shivpuja
Yeah. And they hold people accountable. See, that's the other thing. They're like not only, hey, bring me the data, but also, hey, get me the data. They're challenging us constantly saying, hey, when can you get us this data? We want to do that. When can you get us that data, when can you do that kind of thing. That accountability and that push is also very critical.

Dayle Hall
Yeah. So you talked a little bit about some resources are shared across from this central location. How do you- if you're in this, the merchandising function, you have counterparts in the other functions, how do you work with them, share best practices from that business data, all that business data perspective, which isn't like some of the data science stuff necessarily? How do you work cross functionally with those other teams?

Amit Shivpuja
One is we push a lot of feedback and learnings up to those central teams so that they can incorporate it into those central best practices, knowledge, resources, etc., if we figure out a pattern of doing something. Like we have a new policy for access control, for example, let's say, we push it up saying that, hey, this is what we think will work in a part of the thing. And I think it's replicatable elsewhere in the organization, right? Because people are facing similar challenges. Maybe the ecosystem is a little different, but principally, the challenge is the same.

The other thing that we do is Walmart encourages a lot of cross collaboration where we, peers, meet. So I'll meet my data in a data governance and data product and data strategy peer on the finance side, my peer who's the director of stewardship will meet his peer on the marketing side and the like. But also, the way we operate is that way. So let's say, for example, merchandising wants finance data, business comes and talks to us as the data enablement team, and we go and talk to our peers on that department, and we get the data. Business doesn't need to go to the other side and do it.

What happens then is not only are we identifying the best and the latest and all of that, but we also then have a connection established on how to take care of the data through the flow. We know who are the experts on that side, they know who are the experts on the side and they know why we want this data as well. For example, one of the key things is data contracts. That's what we're working on right now is, how do you establish that transparent relationship that, hey, your table is driving this KPI for merchandising. If we need help, please help us prioritize this.

Dayle Hall
Interesting. Do you ever get moments where you have to push back on some of these requests? Maybe you don't have the data, maybe it's not currently somewhere that you can easily access. How do you- again, we talked about this with lines of business, sometimes they feel like they don't get the support. So how at some point, when either you don't or can't get the data, or you don't currently track it, how do you handle that with the lines of business?

Amit Shivpuja
So it's, again, a business value conversation. So maybe then, let's say, for example, we’re in a scenario, we needed to track certain user behavior data, right? And that enables certain decisions. So what we do is we collate all of that as if it's a business case that, hey, this is what we need, this is what it will do, this is the KPI we measure, this is the value we see, these are the challenges we have come across, that we'll get to overcome with the same. And then we go with our stakeholders to the other team. And let's say, in this case, marketing or the guys who maintain our website. And we’ll say, hey, this is what we're trying to do, this is what we can do, what we want to do. And we then ask them saying, hey, what do you have today, what can you give us today that will get us there, and how can we help you prioritize this into your roadmap so that we can get this in the near future exactly how we want it. So it becomes that kind of a collaboration in order to make this happen, which is why upfront, also I alluded to the fact that it's a very collaborative culture. We're doing this on a day-to-day basis.

Dayle Hall
Yeah, no, that's good. Let me ask you a question because we talked a little bit about value of data working across teams. In your role, how do you- what is a measure of success for data governance-type roles of maybe, Shiv, in Walmart? But if I said, how are you held accountable to success in your organization, what does success look like for you?

Amit Shivpuja
I'll break it up into two pieces. One is the metrics, the KPI, the KR, key result approach. But I'll give you my own opinion after that as well. The KR stuff is productivity, time to get to data quality, the value impact that comes off the data that you've delivered because we capture business value on a regular basis, right? All of those are the KRs. And consistency is another big thing for us. How do you ensure that the same answer comes irrespective of which tool you picked up the metric from, right? So those are the KRs. But I think the simple way of looking at this is, if the business says that, hey, we cannot do these projects or these tasks without this team helping us, we have done our job. Because then we have built the trust, we're delivering to them consistently, and they're getting the value out of it in a timely fashion.

Dayle Hall
Yeah, because basically, what they're saying is they will have- you will have built up that level of trust and relationship that they've seen the success, new ideas, new ventures, something, by working with you. So in the future, they wouldn't want to go headlong into a new project or a new opportunity without knowing that you had their back on the data. Nice, yeah. That would be, I think for any function, if you're seeing a trusted adviser to the business, that's a perfect position.

We've covered a lot of topics. I have two final questions for you. One is, if someone is out there listening to this, maybe they- as I said, chances are, they're not going to be as large as Walmart with many resources, potentially. But you've talked about some good areas to think around particularly the culture. Are there some basics for any organization that, look, if you're going to set up a successful data governance-type model or theme to support the business, what are some of the fundamentals that you would say, look, this has to happen in any organization?

Amit Shivpuja
So let me answer that in a couple of perspectives. One, I think for any successful organization, I call this my tripod model, you need three legs for data. You need a data function, you need an analytics function and you need a governance and strategy function. It's like the camera tripod, right? You need those three legs. So the data function is responsible in one sentence to get you quality, reliable, repeatable data. The analytics function, whether you call them data scientists, analysts, whatever you want to call them, are responsible to pick up that data and translate it into business outcomes. And the governance and strategy piece is responsible to make sure this whole thing happens with a long-term vision in mind. So that's one perspective.

The second one is, many of the things I'm talking about are not expensive or hard to do if you do them early enough. So let me give you one quick one- I'll give you two. One is, today, if you look at even a start-up, let's say they're a 10-person start-up, right? They've just started the product, built the first version of the product, it's creating data. You just need to capture three things, what is the data for, how is it defined, where is it stored? If you start with that basic documentation, you're already way ahead of many companies in terms of data stewardship and data governance, right?

The second thing, or the third one, in this case, is that try and get the businesspeople and the people who are generating data to collaborate as much as possible. So one quick way to do that is before anything is built, try and document what are the success criteria even in terms of data and metrics. For example, let's say you're going to build a new webpage, or you're going to build a new feature, how do you know that feature was successful? Is it number of visits, is it number of clicks, is it number- is it some formula that does that? And educate the engineering team on why that is important, then they will actually engineer it for you. Now, if you do this before the product is actually built, then you're already way ahead in the pool because you now know what you're building towards.

Dayle Hall
And I think that, as you said, if you’re really going to be that trusted advisor to the lines of- to those functions, the early you bring them in, I think a lot of people would be surprised at how much they can be a partner. Because I think they do. I think a lot of people in these types of functions, whether its core IT or in a data science group, actually, they love what they do, but they do want to have a business, like that’s what I want to be famous for. I don't want to be famous for helping the business. So I like that.

Last question. Obviously, there's a lot of hot topics around generative AI and [a lot of] developments. We’re seeing it everywhere, SnapLogic is no exception, where we have some [people] that focuses on that. Is there something specific around either is it- whether it's generative AI, or in the future of data, data capture, data science, data governance, that you're excited about, something that you think, this is going to be a game changer over the next two, five years? What are you excited about in your role.

Amit Shivpuja
Let me put it this way, right? Let me put it as automation. And what do I mean by automation primarily is the amount of efficiency that can be generated using the right tools. For example, generative AI, it's amazing to automate even like Microsoft's Copilot, you can automate code, you don't have to write code from scratch. Or even documenting things, documentation is a tedious task. But if you could document things in an automated fashion, you're so much more ahead. But the other thing that it does is, you would have heard people talk about, oh, we spend 80% of our time manipulating the data, we only have 20% time to do analysis. If any automation, whether it's AutoML or a GenAI or LLMs or whatever, can help you accelerate that, that means you're spending more time doing something with the data rather than manipulating and getting the data that you want.

So that's what I mean by automation, which is, I think in- and I get flack for this all the time is in an ideal utopia, we should not exist as data people. We exist today because there are inefficiencies, but all of our goals is to minimize those inefficiencies, and that's what's getting automated today. Whether it's augmented analytics, as Gartner loves to call it, which is creating dashboards and charts automatically based on the data, or using AutoML models for exceptions and outlier analysis, or data factoring automation, right? Building you the right datasets to run your model with and things like that, is we are moving away from having to put in a lot of effort on playing with the data to using the data more. And that's what excites me the most because that makes us a lot closer to our stakeholders, our business teams and all of that. We are spending more time talking about, hey, what challenges are you trying to overcome versus, oh, should I do this, or should I do that step, or should I write a code here, and should I maintain that and all of those?

Dayle Hall
Yeah. It's interesting what you said about documentation because something that we just launched, which is our generative AI piece called SnapGPT, one of the key features that we've heard from customers and prospects is, you can take a data pipeline, you can put it into the tool, and it will tell you what it does and it will document it. One of the things when we talk to prospects that have 1,000 to 2,000 data pipelines, and then fully- like you said, it [inaudible] the organization, no idea what that does. And they just sit out there running until something goes wrong. So that's the kind of thing. It's one of those little tiny things, but being able to document that stuff by putting it into the tool, and it'll tell you what it does, is I think it's a really exciting time to be in [this business].

Amit Shivpuja
Sorry, I'll just add one last thing, right? One of the things I'm seeing as a trend is I think people are losing sight of what's important. The important thing is not data science. The important thing is not AI or machine learning algorithms. The important thing is being able to apply these things to the business need. If you think about it, there are people who are talking about GenAI and offering GenAI and all of that. But if it doesn't- like we discussed upfront in the beginning, unless it adds that business value, what difference does it make?

Dayle Hall
I think that is the quintessential quote to end this podcast.

Amit Shivpuja
Okay.

Dayle Hall
It’s not adding value, what's the point in the first place? Shiv, thank you so much for being part of the podcast.

Amit Shivpuja
Thank you.

Dayle Hall
It was great to have you. Thanks, everyone out there, for listening, and we'll see you on the next episode.