The Data Cloud Podcast

Making Your Data Delicious with Mani Gopalakrishnan, VP of Digital Transformation at The Kraft Heinz Company

Episode Summary

This episode features an interview with Mani Gopalakrishnan, VP of Digital Transformation at The Kraft Heinz Company. In this episode, Mani talks about turning technical skills into commodities, how to digitally transform your company, and much more.

Episode Notes

This episode features an interview with Mani Gopalakrishnan, VP of Digital Transformation at The Kraft Heinz Company.

In this episode, Mani talks about turning technical skills into commodities, how to digitally transform your company, and much more.

--------

How you approach data will define what’s possible for your organization. Data engineers, data scientists, application developers, and a host of other data professionals who depend on the Snowflake Data Cloud continue to thrive thanks to a decade of technology breakthroughs. But that journey is only the beginning.

Attend Snowflake Summit 2023 in Las Vegas June 26-29 to learn how to access, build, and monetize data, tools, models, and applications in ways that were previously unimaginable. Enable seamless alignment and collaboration across these crucial functions in the Data Cloud to transform nearly every aspect of your organization.
Learn more and register at www.snowflake.com/summit

Episode Transcription

 

[00:00:00] Steve Hamm: So Mani is good to have you on the podcast. And it's particularly pleasant for me because I'm from Pittsburgh and we got Heinz. So it was a big part of my childhood, not just, you know, for snack foods and catch-up , but, but , uh, you know, the Steelers and things like that. So it's , uh, it's an important landmark in my life.

So, you know, our listeners are familiar with many of the Kraft Heinz brands, you know, starting with Mac and cheese, the ketchup, all those kinds of things, but they may not have a broader understanding of the company. So if you'd start by describing the main dimensions of the business,

Mani Gopalakrish: Steve. Nice to meet you. Um, Our purpose is to make life delicious. Really simply stated our business is all about bringing delicious food to people via a variety of channels, such as retail stores or e-commerce. And think about this as wherever you want to buy the food we want to be there. And the most important dimension of our business is we put our heart and soul into making these delicious food products.

And you can [00:01:00] see that manifest in the form of iconic products like Heinz ketchup, Kraft, Mac, and cheese, Philadelphia cream, cheese , uh, Lunchable snacks , and, and, and a lot more of them. Yeah. And again, we're very proud of our purpose of making people's lives. Delicious.

Steve Hamm: Yeah. Now, so many companies have launched internal transformation projects in recent years. Often it involves a combination of things. It data, business models, and even cultural changes. Why do you think this is happening?

Mani Gopalakrish: Look , I, I think if you take a step back , right, the digital transformation journey began maybe two, three years ago. And almost every company got down on this path, knowing that they have to harness the power of data and technology to bring productivity to their businesses or bring growth to their businesses.

But something called COVID happened in 2020. [00:02:00] And in many ways it accelerated the transformation journey by roughly four to five years. I vividly remembering, I vividly remember. I vividly remember sitting down with some of our customers and having a conversation where it is like, yeah, this is a three-year target for your targeting.

Two weeks in, it would be almost two weeks since the lockdown began. We almost like we very well don't pass those metrics that we thought we would hit in four years on, on digital adoption. And so that's one of them reasons, many companies that have, that had already started on this journey are beginning to continue to accelerate this journey for us.

In September last year, we outlined a plan for our future. As a company, we launched a new operating model, vision and values and set for a bold transformation agenda. The end goal is to create an agile consumer driven culture within the company. We believe this combined [00:03:00] with our scale will be the secret sauce , uh, to win in the marketplace. But bring the journey, right. Requires us to learn new skills, build new mindsets, bring in new talent, establish new ways of working and more. We, this is not an evolution. It's a revolution. We call this the digital revolution and we build this digital revolution as a catalyst and a key enabler of our transformation journey.

And we've set ourselves an ambitious goal. We've set ourselves an ambitious goal of being a digitally powered agile enterprise by 2025.

Steve Hamm: you know, often companies kind of like say, Oh, they're going to have a, a technology transformation or they say, they're going to have a cultural transformation, or they're going to have a business model transformation. But basically with your company , it's, it's all wound up into one thing. And that, I think you said digitally powered agile enterprise by 2025.

If you could describe, you know, a little bit [00:04:00] more about the initiative, what are the steps and what's your role in it? Yeah.

Mani Gopalakrish: Yeah, look , I, I think when people start, almost everyone starts with one aspect of the transformation and invariably recognizes that this is all intertwined together. And we are leaning in saying all this is interim intertwined together, which I believe is a, is a, is a smart move on our side. Uh, as we think about the transformation, I think I'd like to zoom out a little bit.

And, and give an analog in the real world, right? If you, if you think about the nineties or the two thousands, most of the people printed out maps and quite two people probably to navigate in a car, right? One person who is driving, looking at the exits and the other person reading out, take right here, take it out here.

Many ways. Most of the traditional companies are there, right? They have a lot of content, a lot of information sitting in Excel, spreadsheets, [00:05:00] PowerPoints , and, and were documents. In that sense, they have digitized their businesses, but still the decision-making begins to happen in innovate like two or three people get in a room and figure out where they need to go. That's kind of the, that's what I call us the digitized enterprise. Take that leap forward. And today, right? I put in the destination, I sit in my car and I'm driving. And as if I veer off of, in my lane, there's a little beep as I'm backing, there's a little bit. If there is someone at the back that the system is giving me insights and steering me in a certain direction, I call this the concept of augmented intelligence and augmented enterprises take that even forward.

We're beginning to see the emergence of autonomous cars, right? Where the cars are doing bulk of the work for you, probably sitting behind the steering and, and acting on it. When you really need to act [00:06:00] this, I called the autonomous enterprise. And in a way, this is the world is going in five, 10 years. And we all have our stuff starting point for us, as we think about where we are and.

How do we prepare for that journey? It starts today. We're putting all of the core tenants into place. Now it began by first identifying all the business challenges and the hypothesis of how we will solve them and the value that it will unlock. So why we are doing transformation, we're still focused on the business value. This gave us our strategy. Essentially. We're now actively recruiting new talent base for the company, such as data scientists, machine learning, engineers, UX designers, full stack developers, scaled agile and experts, and more. And we're putting together several enterprise wide learning programs, such as design thinking to up-skill our entire talent. [00:07:00] We want to crawl, walk and run in this journey and make this transformation. Be so successful that it becomes a benchmark of how other companies need to transform themselves. My role in the company is to lead the U S digital factory that is solely focused on bringing this revolution into the company.

Steve Hamm: no, I think that's , uh, that was quite a big vision and very impressive. Um, 

yeah. Yeah. You know, you mentioned the augmented enterprise and I think a key part of the augmented enterprise is augmented intelligence. And when people talk about that, they're usually talking about combining the collective intelligence of humans with AI and really, you know, so you can have the best of what humans can bring and the best of what the AI can bring and come up with solutions that neither could do as well on their own.

If you could talk about how you're using AI along with the human intelligence in your [00:08:00] organization, I think that'd be really helpful.

Mani Gopalakrish: Yeah. Look when you unleash the collective intellects of humans against the challenge and empower them, magic begins to happen. We are in the early days of our transformation journey and I already see a sliver. I already see a sliver of that augmented intelligence at play. For example. Our data scientists working alongside our functional leads in supply chain identified various key challenges that we needed to solve for one of them was how do we better predicate our inventory thresholds and inventory recovery dates and thresholds, and improve our service delivery to our customers, customers.

In our case, being the retail partners, they basically ingested a lot of data that we received from our retailers. Um, the team ingested it, the team basically ingested the lotta retail data that we get. We also ingested a lot of data that we already have in as a company and [00:09:00] began to say what elements of that data can.

Help us better predict our inventory recovery and how we should update our safety stocks so that we could improve our service delivery. And then we brought in the data scientist to say, how do we solve for this problem? And the data scientists looked at it and said, you don't have three different models.

They ran three different models and triangulated data and the predictions across these three different models and came back and gave us very specific guidance on what we did you get to do. the functional need took that information, went ahead and tested it on one particular skill and was able to prove that this actually works.

You can take the data, you can run a different models on the data and you can create an insight, but insight can then be [00:10:00] applied. To improve , uh, solve for a specific business problem. And look, th th the moment is thinking about, about this particular example and it's near and dear to me is we did this in about eight to 10 weeks.

Nobody asked the team to solve for it this way, but when we empowered them to think about these things differently, this is what they did. It came up. But right now we're training the machine to work alongside us and get better with such prediction. And we are now looking at what is the accuracy of that model, and we're trying to improve it.

And roughly as the module, as being just more data, the accuracy of that model is going to improve and it's going to help make decisions and even better, it can make the decision for us. At some point when it actually alerts the person saying, Hey, there's a problem here. You have to go solve for

Steve Hamm: right, right. Oh, that's great. Yeah. That's it really becomes interactive then I think that's a lot of powers there. So, uh, earlier in your career, I know you did a [00:11:00] lot of work in corporate education programs. And do you see these taking on even more importance in the era of cloud and COVID

Mani Gopalakrish: Of course, there are enough stories about technology powered companies, making traditional companies obsolete. We'd sell them, talk about the people and the culture that kind of failed to foresee that, right? The difference in most of these cases is the ability to learn continuously and a culture to adapt to the, to the changes that are happening in the world.

Quick anecdote in 1999, when I first wrote my, the first piece of code that I ever wrote, and I got paid. Wants to create learning programs that streamed over a 56.6 kbps modem. That's what I wrote in, in 1999. But today the last learning platform I [00:12:00] created took all the content from the world and personalized it for people based on their skills.

It was a kind of Amazon meets Facebook for learning content. A little bit from that journey to this journey. There's plethora of content that's available there. And in building that story is my personal journey. As a guy who started figuring out how to make most of the multimedia work on, on internet to a guy who had to adapt and learn how natural language processing could help us better personalize content for people, how AI and data science can help better contextualize.

Uh, learning for that people. But look, that is a very specific example in it's a powerful example for me to say that's happening in every single domain, take marketing, it's getting more, it's about more personalized [00:13:00] marketing and the way we do marketing is changed. It's no more a game off how good my creative is.

It's a game of my Google SEO keywords and how those keywords begin to work. And that kind of spawns into a whole level of skillset that people need to learn. Take our supply chain example. It is not about me looking at mounds and mounds of spreadsheets and making decisions. It's about training my system to help make the decision for me or give me the other, all of that is completely different skillsets that people need to build. And so to me, even that skill sets are rapidly changing. And I always say this right? Every skill starts out being niche, and then it slowly becomes a commodity. And then you will begin to see a new niche , skill, skill wall that is niche. And it begins to be moderately. If humans don't stay ahead of that curve, they [00:14:00] are first going to become commodity and bend.

They're going to dig themselves out. So to me, learning is the single largest thing. That's gonna keep humanity alive and help people earn their livelihood for long, long, long time to come.

Steve Hamm: Yeah, I think that that will be key because if you, if you, if you don't change, you're going to be, you might be swept aside by, by modern technologies. That's absolutely sure. You know, it's interesting to me, when I think about corporate education and these transformation projects that we've talked about, you know, The it's the most successful companies that sometimes have the hardest times changing because their, their success is so ingrained into the business processes and the culture and just the whole mindset of the company.

So how do you take, you know, even though this isn't your job, I understand, but how do you, how do you deal with a company like Kraft, Heinz, which is so successful, so dominant, so many markets, how do you change the mindsets and the culture [00:15:00] in a place like that? And specifically in the it part

Mani Gopalakrish: So first and foremost, we are blessed with the leadership that recognizes the need to change. And to me that makes the entire conversation easy. I wouldn't say easier makes it easy. Right. Then comes the next part, which is how do you rally the entire employee base to recognize that the good part of that is we are living in a world where we actually don't need to do much.

The system itself is pushing everybody to understand, empathize and change. So, so to me, like I look at it and say that are. Change agents and change instigators everywhere in the industry, within the company with not leadership and all the fonts are playing a pivotal role [00:16:00] in pushing each other, pushing each other to get to a better outcome as it.

And part of that is also how we evolve our, our it, right. The way I think about this. And again, I like to see, I like to draw analogs if you haven't caught on to that. Uh, in the 1990s, when Microsoft word came along, right? And Microsoft Excel came along and the vision to put a desk, a computer in every desk began to come to life. The power of content got democratized now in companies just about everybody can create content. Where does that journey taking us in the next level? Right? Like you take the journey to the next low code, no code and food code apps. Almost everybody can become an app developer. How does it, but think about its transformation in this particular scenario? I mean, nobody needs to push anyone here. The system [00:17:00] is pushing everyone to work differently. And so, so I, I do think that bringing something to something like this to life requires leadership buy-in and recognition, which we have. We need smart talent who understands the vulnerability and the need to change, which we have. We have the system that is, and when I say systems, not tech system, but the overall processes that needs to push each other , um, to, to evolve and change. If all of those things are in place, what's the needed is the focus. And that focus is what we, as a company are putting through this digital revolution to accelerate our transformation.

Steve Hamm: now one of the great enablers of this kind of transformation, both the corporate and the it. Is the, the rise of the data cloud. And so I wanted to drill down a little bit with you, [00:18:00] and if you could tell the story of your, of the company's data cloud journey, when did that start and why

Mani Gopalakrish: Look, a lot of my colleagues have played a pivotal role in helping us get set up here. The journey actually started in summer of 2019. It started with the broader transformation of, of our moment from on-prem to cloud. And that's how it got started. And. Why do we look at everything that we need to transform into the cloud?

One of the big elements was data, and we began our journey towards the data cloud transformation in about 2019 summer of 2019.

Steve Hamm: and specifically , um, snowflake. Uh, what's the role of snowflake technology in your data cloud strategy and how are you using it and what benefits are you getting

Mani Gopalakrish: Yeah. Snowflake is our company-wide the data repository or [00:19:00] data warehouse. We have built several pipelines and now have migrated trillions of records into snowflake. From there we're beginning to generate the intelligence needed to run our business. We've also tapped into snowflakes, a data marketplace to accelerate our needs.

And the real benefit here is agility, right? The speed with which we can move Walliams of data, structure them. Build intelligence and democratized data-driven decision-making that is so fundamental to become an intelligent enterprise. And snowflake is that intelligence enabler for us in this journey.

Steve Hamm: yeah, yeah. Now I know there are a couple of key elements of flex technologies or attributes that are important to you. Uh, one I understand is cloud independence. Tell us about that.

Mani Gopalakrish: So we were going to be putting all our data into snowflake. And one of the things we [00:20:00] wanted to make sure was we stayed independent of a very specific cloud provider. There are many reasons for it, right? One is just the business prudence of it, I guess, is the starting point. So we looked at, we looked at snowflake and said, look, it can run on Google.

Can crowns run on Azure? It can run on AWS. And we had the flexibility to move the snowflake and Waterman. Should we ever, ever need to switch our cloud provider? And I hope that's not the case. However , uh, we needed that freedom and that empowerment so that we were not saddled with a provider for something as strategic as data.

So it becomes a very important concept for us.

Steve Hamm: yeah, they could put really have some leverage over you, but if you're independent and w or even multi-cloud , uh, and you can, and you're flexible, you can really, you call the shots basically. Right?

Mani Gopalakrish: Yeah. Yeah. And it's less about. [00:21:00] Calling the shots. I believe it's more about de-risking premonition technology standpoint , the, the investments and having the prudence and, and, and moving towards more of a multi-cloud hybrid cloud like environment. But again, the starting point needs to be cloud independence where we can, and that permeates in the way we build our software applications through Kubernetes that permeates in our decision-making on, on cloud , uh, data, data cloud , uh, with snowflake.

And that permeates our decision-making on how we build our software.

Steve Hamm: Yeah. Now the S the snowflakes decision to separate storage and compute was one of the first decisions they made. And my God, it turned out to be very strategic. And I understand that this is one of the things that's valuable for you. So if you could talk about why is that so important to you?

Mani Gopalakrish: Yeah. If we, if I go back again to, to earlier parts of flight, right? [00:22:00] One of the never ending discussions that we would have when we were building software is how much volume of data's come in. How many transactions are you going to put in? And how did the database locks work? This is the kind of discussions that at least I grew up on in my early career.

Right. But when I look at the separation of storage and compute, I sit here and say, why wasn't this, the case when some of the other databases were built earlier on, right? And I give you an example of how this , uh, helped us, right. We had to migrate. Half a trillion records from another large data warehouse into snowflake.

And we were able to accomplish that roughly in two sprints, less than four weeks, because the compute and storage were separate. We were able to scale up and scale down the compute on an as needed basis and push all this data into the, into the other [00:23:00] system, into the snowflake system, without having to overtly worry about the performance implication. So, so that that's one place where, where it actually helped us and it continues to help us. Uh, and every time we need to amp up compute and amp down compute, we're able to make sure that that we're not , uh, degrading the performance of the storage side of the house.

Steve Hamm: I know that another main attribute of the data cloud is the ability to share data much more easily, either within an organization or between organizations or, or even , uh, getting data from third parties. And I, I believe , uh, I understand that you're looking at the snowflake data marketplace as a strategic element of your strategy.

So it'd be great if you could talk about that as well.

Mani Gopalakrish: And th the way I look at it is data. Marketplace is probably one of the smartest males. And smartest movements that's occurring in the industry. [00:24:00] Tactically, the way I look at it is there's commodity data, right? It's your econometrics data, weather, data, and other such data in the absence off a marketplace would have been probably a few sprints of work from a data engineer to make sure that the pipe, the pipe, the data pipelines appropriately, the, it was sized for , uh, the, the right load and so on and so forth.

Right. But with the data marketplace, it's a click and a credit card away, right. Just go in and put a click and get the data in very similarly. The we did this concept can be extended to a lot of first, second, third data party. And how do we bring all of that? In-house how do we automate all of that? So our focus is really not on the plumbing, but on the insight generation and data marketplace [00:25:00] helps us do this in a, in a powerful way.

And I will actually go as far as saying that a lot of our, and I like to call this with all due respect, dumpster fire challenges happen when a, when a data marketplace doesn't exist. And in those integration points is where we have a lot more challenges,

Steve Hamm: so a dumpster fire is when you need to integrate data quickly, but you don't already have the pipelines ready. So you can't, you can't add a grade, but if you have the marketplace, it's already, all the pipes are there and available and all the data is there and available.

Mani Gopalakrish: Yeah. I mean, I think there are a couple of levels of Doubleclicks onto it, right? The first DoubleClick would be what you just said, right. I need a data really quickly and this data is easily available, but I can get it into my environment because now I have to do a security review and I now need to do the pipeline building.

I need to now size up the performance. I [00:26:00] then need to clean up the data. I then need to bring it up into my snowflake environment and then I need to profile it. Like that's one level of data, man. It's just fighting to say it. Imagine putting the people to work on easy commodity data that should be available.

The second level of problem there is. Holiday integration points. Not every company is sophisticated and they're all working on getting there, but it's just, it's just a grind, right? When we get data, sometimes it's in a spreadsheet. Sometimes it's on a portal that you need to log in and grab the data. And sometimes you have to write a robotic process to log in and pick up the data on a certain date and time.

And now you need to set up monitoring for this robotic process. Then you need to set up monitoring to make sure the data is actually there. And that particular portal is not down. This mode of [00:27:00] complicated data exchanges needs a permanent solve and, and data marketplace can actually help alleviate a lot of those problems.

Steve Hamm: Hey, do you find that you're just using a lot more third-party data? I can get much more of that contextual information than in the past, just because it's available on the marketplace. 

Mani Gopalakrish: Yes. We're using a lot of data from the data marketplace for sure.

However, there are still a few things that we as a community need to solve for the first problem is that overall lack of that ecosystem maturity, which is the data interchange standards gave you an example of how data is exchanged across a variety of different formats. And each format requires a specialized approach on how you monitor and make sure the data comes in.

The second thing is the mindset. It took a long [00:28:00] time for humanity to understand is we live in this complicated, intertwined economy where all of us went together. And if even one part of that, that doesn't work well, it impacts impacts the other. And what are the same place with data? Holding information power for sure, but shaving, it makes you more powerful.

Right, because now you are not only getting the value of your data generated in your ecosystem, but you're also getting value out of the data generated in the other ecosystem. And so we somehow need to get over that mindset and how that manifests itself is in this complicated terms and conditions of what you can, you can do and not do with the data you get.

And sometimes I feel like the food's on the table, but you can't heat it because of a contractual term and condition of how the data was agreed upon. So while the data marketplace is a [00:29:00] great first revolutionary step to solve for this problem, there are other things around the ecosystem that needs to make sure for this to work in a more effective, and we're all getting there.

And I guess the first stepping stone is the marketplace.

Steve Hamm: Yeah. Yeah. Interesting. Now 

great. Now, whenever people are talking about it and changes and new technologies, they always want to hear about the metrics, the numbers, you know, prove it really worked or show us that kind of thing. So it would be great. If you could talk about, you know, in a, in a quantitative way, how snowflake has helped you

Mani Gopalakrish: Yeah. We decommissioned our on-prem data lakes in nine months. We moved in that journey , um, which otherwise would take probably multiple years in, in nine months we have 500 terabytes of data and growing and mind you, we're continuing to pipe streaming data into [00:30:00] snowflake. And by the time we finished this podcast and by the time it's probably published that number might have shifted. And as I, as I said , the, the snowflakes ability to separate compute and storage helped us migrate half a trillion records in roughly.

Steve Hamm: sprints of about four weeks. That's really incredible. Yeah. No, that the speed with which people can get stuff done these days is just remarkable. When you think about it. I mean, I covered when I was at this, this week, I covered enterprise software, enterprise hardware in like the mid nineties and man, I mean, people spent years putting together an SAP, you know, program a corporate wide and, and hundreds of millions of dollars, but things have changed dramatically.

And I think a lot of it's due to the cloud. Yeah. Um, I w you know , we, we, you talked a little bit about challenges a minute ago, but I want to go back to that question, because I think even though. The data [00:31:00] cloud and the cloud make a lot of things easier. It's not really, totally easy yet. So what challenges have you faced as you migrate your data and your analytics to the cloud and how have you overcome them?

Mani Gopalakrish: Yeah. I almost break that down into strategic issues and tactical issues, but tactically tech is not the problem. We can put the pipelines, start plumbing, the data, however, letting everyone know what we have and how we can use them with confidence. Is an ongoing challenge. That challenge gets amplified as you take the big data and start making it usable at a micro level via KPIs, that aspect of validation and accuracy, and making sure that the KPIs you're seeing when you're making very critical, precise business decision is still human dependent.

And so there's an element of [00:32:00] training. There is an element of maturity, and then there is an element of effort that goes into solving that problem. And as it stands today , right, we're doing all the right things like profiling data , um, using AI to validate and any anomalies in the data. And, and we're trying to automate that part of the journey, but there is still the last mile of humans that that is a heavy lift.

So I think tactically that's one big challenge strategically. I think there are, there are a few different things. And as I said, data, marketplace is probably a good first step in revolutionizing and solving this problem, but there are broadly three different things that still need to be solved for first one is overall ecosystem maturity, which is there's there's lack of data interchange standards.

We still continue to receive data from portal that somebody has to log in and download on Monday morning at nine [00:33:00] o'clock. Yeah, we can put robotic process automation into it, but now that's another failure mode that we have introduced into the entire entire conversation. And I expand that too. You got FTPs.

You got SFTPs, you got emails, you got spreadsheet. So that lack of data interchange standards across the enterprise is still mind boggling and it's probably problematic. The second thing is the overall mindset of how we can exchange data between enterprises. And we're still learning as a, as a, as a ecosystem of enterprises.

What data should we keep? Because it is proprietary. And what data should we share more freely because we can generate value across two ecosystems. And across two partners, as opposed to keeping the data and holding the data, which then limits my ability to generate value out of the data. And the third one.

[00:34:00] And this is how all of this manifest, right? That, that lack of overall ecosystem maturity and the mindset of holding information versus sharing manifests itself in complicated terms and conditions on data contracts. Sometimes you feel like you have food on the table because you can, but you can't eat it because there is a contract that says you can see the data, but you can use it for this purpose.

And, and believe me, this happens pretty much every other week. I would see some contractual clause to limiting us from doing something. And so to us right now that is a grind, but like with every company we need to go through and kind of get our arms around it. However with the, hopefully the maturity of the data, marketplaces and maturity of these , um, of these conversations, we will begin to see where data can be a click away truly.

Steve Hamm: Yeah. Yeah. It seems like, you [00:35:00] know, the more programmatic you make things, the easier it will be to deal with that. I mean, those, those special terms and conditions, if you make it programmatic than a human, doesn't have to all read through the list of what we can use and how we can use it. Basically, when you try to use something in a certain way, you either, you either get it or you don't, or you get it delivered in the way that it's intended.

So it's yeah,

Mani Gopalakrish: Yeah. And I think, again , it's, it's all a matter of maturity. It's a matter of need that not only we as scrap times have to go through, but our partners also have to go through and evolve.

Steve Hamm: yeah. Yeah. Now Kraft times sells its products primarily through its customers, which is retailers and distributors in order to reach the end customer. So how does the data cloud help you share information with your business partners? So you can better meet the needs of consumers [00:36:00] and manage inventories, you know, both the business end and the demand end

Mani Gopalakrish: Yeah. I mean, I break this data down into multiple different facets. First is commodity data that's right now, a click away with a snowflake. I don't have to sweat bringing in weather, data, economic metrics, data, and other such information that I need to augment for my , uh, day-to-day , uh, research, our day-to-day models and so on and so forth like that, that I think has been drastically simplified.

We have. Partnership with a lot of our retailers where we do receive data from them. And that data is, is a mix of custom data pipelines and a mix of , uh, potentially marketplace. So we're, we're continuing, that's one area where we would love to see a more industry wide partnership to automate and make this process simple because all of us can collectively win in this ecosystem.

[00:37:00] And then we continue to augment a lot of third-party data into the, into the system and, and that third-party data is still not easy and that probably needs to organize itself around the marketplace before we make it becomes easy for us , um, along the way. So, so I think that's almost all about data ingestion internally for us to distribute the data.

Snowflake is help. Us in that distribution in a very unique way, in the sense that we're able to now take the big data, the macro data, and create a unique data architecture where we can send the small snippets of data to various business units. We launched , uh, a learning program , uh, a few colleagues of mine launched a learning program called Zenit, where they train these citizen developers who can create insights out of that data and then accelerate, help businesses [00:38:00] accelerate.

Decision-making. Now all of this would not have been possible without some of the data cloud infrastructure. So that's how we begin to see the power of data cloud, help us with exchange scalability and distribution of content.

Steve Hamm: Yeah, no, your company is going through a major transformation and actually within it, there are kind of sub transformations. And from a previous conversation, we had, you referred to one of these by the highly technical term of consumer obsession. So Ron I, and the way I understand this is rather than focusing, mainly on marketing your individual brands, you're seeking to understand the shifting consumer tastes and demand, and you're developing new products or evolving the existing or evolving existing products to meet them.

If you, if you could talk about how data helps you do that,

Mani Gopalakrish:  At the, at the heart of this is us creating a total understanding of our consumers. We're, we're living in this interesting time [00:39:00] where we have a lot of data and that data comes a great responsibility. We take our consumer data and Nona minds and do aggregate analysis on, on that particular data. So, so I may not know that Joe down the street brought our catch-up, but I would know that people in Chicago, Illinois area with such a demographic demographic brought such and such product, that could be our product , our, our, our consumer, our competitor's product.

We are also able to decompose. The ingredient levels in our products. And we are able to not only use that for building enterprise cost drivers, but also Al part research and development teams on how to think about creating new products. So we're able to bring in [00:40:00] consumer three 60 , um, to life by not only our own data, but also working with several partners and bring it to life in a very anonymous and aggregate way.

We are also, and I may use this word more generously called competent or three 60, which is where we are looking at our products, decomposing our products , uh, and then matching it up against our competent or products and saying where and how we can continuously improve our products. Right. And then maybe I will take that even further and say insights three 60 and how we're able to clean.

Everything that's publicly available and identify the demand landscape and the trends even before they are here. Now, all of that are in various stages of maturity in the journey as we speak. And, and , um, I faced the consumer three 60 journey that we have started an accelerated, gives me confidence and hope that we can help.

[00:41:00] We can create a data platform that can help the entire company be more consumer obsessed,

Steve Hamm: right. That makes total sense. Now many, I'm going to ask you to put on your visionary cap for a minute now. So looking out five years or so, what are the big changes that are coming in the data cloud and how will that help transform business and even society?

Mani Gopalakrish: Yeah. So, so when I put , um, my five-year cloud and , uh, sorry, when I put my five-year half. Right. Um, and I'm going to be a little bit more pragmatic here and say, The first and foremost that's going to happen is this maturation of data, marketplace concept. Everything kind of begins to go into a data marketplace and the data interchange is going to become more seamless.

We as a society, we as human beings and we, as companies will begin to understand what data can be shared for. Good. And what [00:42:00] data should not be shared. Um, to prevent any privacy violation. So I see that coming into play in the next five years, I also see the concept of data networks coming into play. So the first step is data marketplaces, but once this market places are alive and kicking, we would probably now have a better understanding of how the inherent connections between data across these marketplaces.

And we're able to bring a network to it. So I'm able to easily strain my inventory beta to my sales data, to my consumer data, to my procurement data. And I'm able to create an enterprise that is completely run by data. And I see that happening in the next five years or so. What are the implications? The implications for that is data has to go from being. Something that's guarded in a [00:43:00] company to becoming a robust platform and becoming a service. So the usage of that can be democratized right now. Data science is a much sought after skill in five years, it's likely to become a commodity. And that doesn't mean that we won't need data scientists, but a lot of the models that nowadays takes a lot of programming time will probably become commodity.

And there'll be like more advanced things that the data scientists might have to build coming out of that. I see tech democratization, this entire no code local food code , um, concept is likely to turn a lot more citizen app developers in the company. Just like how everybody can create, like say Microsoft PowerPoint or word documents.

People will be able to create their own apps and that's going to then generate more data. [00:44:00] And now we are beginning to see how a huge flywheel in effect and all of this is going to have a huge impact of how companies are organized, how companies work together, how people in the company work together, what kind of skillsets are needed, how it needs to transform itself, how businesses need to build their strategy.

I see a huge shake up in the making, partly driven by the trends that I kind of just highlighted.

Steve Hamm: I think that's a really cool idea. Especially the democratization of app developing. I mean the low code, the no code kind of stuff. If, if everybody, part of everybody's job is to be an app creator , uh, you're creating new capabilities for the company, not just like content, like a PowerPoint, but something new that can be done in some new insight that could be received.

So I think that is really cool. You know, at the end of these, these podcasts , we, we typically ask a lighter, more personal question [00:45:00] and I understand that your passion is coding now. Clearly that's not part of your professional life anymore, but you, you love it. And you still do a lot on the weekends. Tell us about that.

Mani Gopalakrish: Okay. I, I told my wife this, when we were first met that I told her you're mad in a workaholic. That's the only thing you need to know about. Um, and, and that's been true in our 10 years of marriage. I love coding. According is my form of creativity. I love creating, creating apps and solutions. So many of them don't see the daylight just partly because of my schedule, but there are times , um, there I have built and launched apps , uh, in the store and they are still alive and kicking , uh, with the fairly healthy usage rate.

I mean, I built , uh, AI Marti, a named entity recognition engine, just for fun, a couple of months back where I took all of the resumes that are out and figured out what job and what skills are hot in the marketplace. [00:46:00] And it was just for fun. I blog , um, personally on business, digital transformation and code snippets.

And, and that's what I do. Um, I I've most recently learned, obviously I've most recently learned how to create a named entity recognition engine. Prior to that, I learned react native. Prior to that, I learnt react JS. Prior to that, I learned like Google cloud and Firebase. And most recently I learned a few elements of Azure because we are, that's where we are beginning to put our, put our , um, making, making it our enterprise cloud.

So to me, I, again, I go back to what I said earlier in the past podcast learning is the single most important thing that's going to differentiate. And my passion gives me the ability to learn that learning helps me to apply that effectively at workplace, and that that helps my company. And then that helps my profession.

And it's a flywheel that I believe has created.

Steve Hamm: Yeah, yeah. It's [00:47:00] interesting how you, you're melding your professional life and your private life and it's all one. And I think that's true for a lot of us. The trick is how to balance the two, but it seems like you've got it. So I want to, I want to thank you so much. It's been a, it's been a really great conversation today.

You have some big ideas about enterprises, which I think, you know, a really go way beyond your industry and way beyond technology. I , uh, I really thought that one of your key insights today that I'll carry away is the idea of the business ecosystem. You know, even within companies or between companies in the past, you know, business leaders tended to think, Oh, I'm going to, I'm going to hoard my data.

That's my source of power. And, and you said, yes, that is power, but they gain more power from sharing it. And I think that's a hard lesson for people to learn, but once they learn it, they and their, and their organizations are going to be more successful for it. So thank you for that insight.

[00:48:00] Mani Gopalakrish: Yeah. And Steve, thank you again. It was, it was a fantastic catching up look. I'm very passionate about , um, the, the. The journey that all these companies need to take to be a more autonomous enterprise. And, and I'm just humbled to be playing my part in it, but crap times, and , uh, I, every day I wake up thanking our leadership for giving me one, the opportunity to share a similar vision there.

They believe this was important for the company that makes life a whole lot easier.

Steve Hamm: thanks,