Podcast #151 New Data Revolution

Milind is the world-recognized expert in designing, developing, and operationalizing some of the largest data platforms, including the Indian elections, where the Modak team processed over 650 million voters. For the past 25 years, he led numerous global Fortune 100 companies in their data-driven business excellence journeys, providing thought leadership solutions delivering their excellence across business strategy, analytics, and modern data platforms. Milind has six patents in data management and analytics and is part of the guest faculty at the Indian School of Business in Hyderabad, India. Milind, welcome to our digital thread podcast.

01:32

Thanks, Ken, happy to be here. And first of all, congratulations on completing 150 podcasts.

Yes. I inferred online that the line-up would be even better, and you're the first to kick that off. And so, congrats to yourself for being in the lead position there. Obviously, with your journey and especially talk a little bit about

But first I always like to start this off by understanding one's digital thread. In other words, the one or more thematic threads that define a digital industry journey. So, what would you consider to be your digital?

02:11

First of all, Ken, that's an interesting question because tan exciting to go back and figure out my digital thread. So, I'll start with my journey and then talk about what my digital thread is.

My journey started 23-years ago, and I think I'm one of the fortunate ones continuing with the passion, which I began about almost 30-years back. I always wanted to be in computer science. What I did is I went ahead and completed my bachelor's in computer science in India, and then I moved to the US for my master's in computer science. I was fortunate to get an internship opportunity during my master's at Microsoft. And then, I again got an incredible opportunity to intern at IBM TJ Watson Research Centre.

And while interning at IBM Research, I was given an incredible opportunity to go and solve a business problem for IBM's consulting arm. They were building one of the largest data warehouses in its time, and I'm talking about 1998 for one of the largest insurance companies in the world. And here, just an intern going there awed by the whole consulting world and helping them solve some of the problems because the research felt that I was good enough to go and do it for them.

That started my whole data journey. I was incredibly fascinated by the world of data and what we were trying to do there at that insurance company back in 1998. We were attempting to build one of the largest data warehouses at this time, and if I'm not wrong, right now, 18 terabytes sound like a small amount of data.

A small amount of data, but back then, our data drives came in two semi-trucks. Then we assembled that whole warehouse on the IBM platform and then started to collate all the data, bring it to a place, and begin churning insights to help that company market. So, if you ask me what my data journey or digital journey has been, my digital journey has been building data platforms at scale and then turning data into an asset for some of the largest enterprises. Of course, I was fortunate enough to work on multiple continents, so I worked obviously in North America, I worked in Asia, and then I spent a significant amount of time in South America, observing these cultures and then how things get done, and leading large teams across different continents is what I would call my digital thread.

Luckily enough, I'm continuing that passion. So even after 23 years, I am still passionate about building large-scale – petabyte-scale, in fact, data platforms. And I'm thrilled that I can accelerate that process for enterprises and our customers, where we are rapidly delivering value from that data.

The other thing I would point out, given where we are right now, is that if you were in the data domain, I would right now state that we are in the golden age of the data - in the midst of what I call new data revolution, which is slowly transforming our lives. And that's what this journey has been. So, I'm looking forward to this continued journey in this space.

It's interesting. So, starting with an internship position, you spent 15 years at IBM and ultimately co-leading their business analytics and optimization practice. You mentioned already a lot of the breadth of experience in terms of countries, clients, use cases, etc.

What were some of the key trends that you observed across that time while there, and relative, of course, to this new data revolution?

06:07

Absolutely. My profession at IBM was gratifying for me. I joined as an intern. As I stated earlier, I became a consultant, and then I eventually became a leader managing a large team. And in that whole process, I got to work with some incredibly creative colleagues, who had probably molded me into what I am today. I always honor that and respect those colleagues who have become lifelong friends now.

And another thing is since I worked in the consulting arm of IBM, it always provided me a unique outsider perspective while I was working with our clients. How their processes work, what makes them tick, what are the different business problems that each one of the companies is trying to solve. If I continue to go back a little bit, right now, the world is what I call caught up in the hype of machine learning and artificial intelligence, rightly so as this is incredibly fascinating.

But in 1998, when I was working for one of the largest insurance companies and joined IBM, I was very fortunate to work with some of the brilliant mathematicians from IBM labs who were hand-coding the neural net algorithms to work on how to optimize marketing. So, it was fascinating for a 21–22-year-old to be exposed to that.

And then from there, in 2005, I was asked to join as a founder-member in a Centre for Business Optimization, which was IBM's initial foray into predictive analytics and machine learning. That eventually evolved and merged into what was known as business analytics and optimization practice. This group was very close-net it was a 15-member founding team. So, I have seen from my perspective how the evolution of what we term, analytics and optimization, has happened during that timeframe, from 2005, and I had a front-row seat for that transformation.

When I started my career, the problems were defined as data warehousing and data marts, then eventually they were called business intelligence. Then the term became analytics, and then soon it became advanced analytics, then then it became Master Data Management, and then suddenly it became Customer 360. Then obviously got into the big data world, and right now, it is termed machine learning and artificial intelligence.

What I feel is fundamental is that we are trying to solve the same problem, which is you are gathering the data, you're managing the data, and when you're processing that data to gain valuable business insight. Whatever the term we use, it's the old wine in the new bottle – and we're repeating the same thing. What I see in that whole journey is around 2007, business leaders started becoming very serious about data, and there are many reasons for that. One is they wanted to gain a competitive edge in the internet era. Hence, businesses started shifting from brick and mortar to this complex, connected world with different touchpoints. They were deluged with data coming from these touchpoints, internet portals or apps, or whatever that is. So, they became earnest about investing in a data asset.

And then, the financial crisis happened, which accelerated the change where data was seen as one of the driving factors, and there were multiple changes associated with that. So, because of these data sources popping up, they were looking at how we leverage this data, how we manage this data and struggling with the deluge of data.

And then suddenly, the world caught up in this hype of big data, and the primary trend I saw, while sitting having the front row seat, is the term which I call a centralization of data. So, what happened, is organizations decided that they would put all their data in one place and then give access to their data scientists. They then thought that this would accelerate the decision-making process or accelerate the mining of information from the data. But what they failed to realize was that the problems were not solved. The problems were basically, how do I identify valuable data in the company; how do I trust my data; how do I know what my data lineage is?

So, the problems were similar, and they were most diverse; if you go back and look and search on this, there are thousands of articles that state that 70% to 80% of these big data projects or these projects fail because they are incredibly hard to build and leverage them.

And another thing that people probably don't realize is that only five to 10% of the data generated by a business is actually useful and that 90% of that data is artifact generated from their applications and business processes, which is not as useful. So, how do you get to that 5% of data, identify it and then figure that out? So, if I look back to my IBM days, the central team was the centralization of this data, and we were not hugely successful at that If I might say so.

You co-founded Modak analytics in 2013 with R D Joshi, also an accomplished data analytics leader. Perhaps you've already spoken to the problem set a moment ago, but what problem did you set out to solve and for whom?

12:09

When I decided to quit IBM, and that was an incredibly hard decision coming from, I didn't have any entrepreneurship background, and my family was not into business. Everybody was employed. And especially given my Indian background, everybody was scolding me that you're doing so well. Why are you now quitting and starting this new enterprise? Do you really need it when your life is all good? But I did make that decision, and the reason I made it is, I wanted to fundamentally make a difference in how we perceive data and what I can bring to the table based on my 15 years of experience at IBM.

During that time, one of the things we started noticing is that companies were hiring CDOs, and those CDOs were tasked with building out data capabilities for these companies. One of the things we noticed is the failure rate, they were trying to build these data platforms, and they were failing very often. The opportunity we saw was that we could significantly reduce the time it takes for organizations to build those data platforms and actually derive value out of them. And, if I might expand on that a little bit, there is a three-step process for turning data into an asset, and the first step is you start gathering the data, then next step you start managing that data, and the third step is you actually start processing the data to what I call into knowledge. The data scientists then pick up that knowledge and turn that into meaningful business insights. So, this is how the process works.

Put as simply as I stated, this journey of turning that data into an asset is incredibly hard. And the reason why it is incredibly hard is that it takes so much time for organizations to build those data platforms and then finally implement a business use case where they are asking those questions.

Generally, what used to happen, is that the businesses at any given point of time have questions, and the way the business is evolving, within six months or a year, those questions change. And what questions I asked about a year to one and a half years back are no longer valid because they have probably done some gut-feeling and went ahead with some decisions. The reason was that it is incredibly important for enterprises to quickly answer those questions using the data, and that's where we set out to build it. And we thought we could build what I call a modern data platform and accelerate that time to value by five to ten X. And the way we set out to do that at Modak was to adopt a different strategy than, let's say, other service providers. So, at Modak, apart from having a talented team of people, we also have invested heavily in developing products and accelerators to achieve that time to value. What we have right now is we have created a product called Nabu, which is a fully-fledged solution in production at some of the largest organizations in the world.

Nabu was built ground-up to solve this problem, it was built up in this cloud era along with different customers, and that is differentiating for that. So, we knew the customer problems, we were working with the customers, and then we started building this product along with them to solve the problem of how we accelerate. The good news is we won accolades for Nabu, and it was also mentioned by Gartner.

Probably I should simplify this a bit. The analogy I will give to help the audience understand what I mean by this acceleration. Let's take the analogy of a home-cooked gourmet meal. What happens there is, I create an ingredient list, I go to the grocery store because I probably don't have all the ingredients in my refrigerator, and then I shop for that, come home, I prep my meats, cut my vegetables, and then I'll probably cook two or three sauces to go along with the meal. Then finally, assemble that meal and visually dress up that meal so that I can present it to my family and friends, and then we all enjoy that meal with a glass of old wine.

And that's what is exactly done by enterprises when they deal with finding data or trying to convert data into an asset. They follow this gourmet meal approach where they are building, every time they are going and building. Now, what we are trying to do is make that into a commercial kitchen for a restaurant. And in the restaurant, the ingredients are pretty much delivered to your doorstep every day. You take the meats, you take the veggies, you prep them, you precook your sauces, and when the restaurant service starts, you can build these known recipes within 15 minutes and deliver that to your customers who are incredibly happy.

So, what we're trying to do with Nadu is accelerate the data operations in a company, from the gourmet meal approach to the commercial kitchen approach. And, if I may point out, I'll give you an example.

In our recent work with one of the largest pharma companies in the globe, we were able to put up a modern data platform in under 12 weeks. We brought in data of 150,000 data sets across the organization, we profiled that data and made it available to the data scientist. This was done in 12 weeks, and immediately after that 12-week period, now that we have our commercial kitchen approach, we started working in parallel on eight different business use cases. That is how we have delivered value to our clients.

If you talk to anyone in the data industry, what I would say is, this is like a massive acceleration versus doing the traditional way. And that's probably why I started or co-founded Modak, and it's been a very fulfilling experience.

Yes. It sounds like it has. And I love your analogy of the commercial kitchen. I might even go so far as to call you the fast food of DataOps! But I'm not sure that would have a positive connotation.

But I understand you guys are already over 350 people at Modak, so quite a journey already. And that at least lately, as we were talking about, e-health applications are driving much of that recent growth. What health care sectors are you focused on, and what are some of the factors contributing to this?

19:17

As you pointed out, that was incredibly rapid growth for us. So, until 2017 we were still a very small organization. At the beginning of 2017, we were just 22 members, and today we have scaled up to be a team of 350 members. However, we still remain to our core ethos, which was, we are going to be a boutique company very focused on bringing some exceptional talent that is devoted just to data engineering. So, we don't do any other thing apart from data engineering. Even though we call ourselves Modak Analytics, the majority of the work – I would say 98% of our work, is focused on the field of data engineering.

We don't even go into analytics because I fundamentally believe that 85 to 90% of the work a company does when it's dealing with data is data preparation and data engineering, and 5% or 10% of the work is devoted to what I call data science. More and more I see when people come to me for advice, I tell them there is AutoML coming, where it has become a computation problem; you throw a data item to the computer, and it will tell you which algorithm is the best suited to this job without needing a human. So why would you want to enter that field, which is going to be automated?

Of course, I might be controversial there, but that's my opinion coming from what I do.

Now, coming back to your question on what is driving our growth, and what are the sectors we're focusing on? Our current focus has been mostly working with some of the top life sciences and healthcare companies. What that requires is, it requires deep domain expertise In that sector and an understanding of the data elements in that sector. What we are seeing very recently, especially after the pandemic and even before the pan is that there is a massive investment from these companies to build out their data infrastructure and developing new capabilities around the data.

And I will tell you what is driving that. Let's take the example of the pandemic. The pandemic, which was an exemplary example of human ingenuity. Across the globe, companies tackled to manufacture a vaccine in record time, without data it would have been impossible. So, for example, the COVID virus, they sequenced it in a matter of under a week, and then the results were published. Then the pharma companies and every researcher took that data and started looking at what to attack, and then they found the spike protein, and then they went after it, and we created a vaccine and hopefully, we as humankind will quickly get out of that pandemic. But that tells you about the impact the data is having on this.

What the business leaders are now pondering is, how can it take the same learnings that have happened in the pandemic, and then accelerate drug development, or reduce the cost of patient care. Or how do I use the data to better monitor and improve patient outcomes? Also, one of the needs which were driving even before the pandemic, and this is actually a social need, the need to reduce the healthcare costs and the cost of the medicines, in the US it's incredibly costly, and in Europe, it has become costly; how do we reduce the cost and improve healthcare. And what we are seeing with all of this is, there is a new data revolution happening in the e-health and healthcare space. What do I mean by that?

So, if you think about what has been happening very recently, this world of IoT, sensors space, and variable technologies are coming into play, and they are sending out lots and lots of data. So, for example, my dad wears a Samsung watch that takes blood pressure readings, and we can monitor his ECG reading because he is a heard patient, and that allows us to take those readings, send them to his doctor to do the remote care and the virtual care. A lot more companies are realizing that these life-threatening chronic diseases can be better managed at home. It will reduce the cost and the data coming from these variable devices actually will help them do more preventative maintenance as far as health is concerned.

What that also means when you look at the companies which are doing this, is there is a constant flow of data available to healthcare professionals, scientists, or whoever wants to use metrics like blood pressure, your heart rate, or your sleep patterns, and they will use this information to diagnose and treat you better.

Another thing that we are seeing in the pandemic is the evolution of remote care. Right now, for some of the major diseases, you probably don't even want to visit a hospital. And in fact, recent research has shown that for some illnesses you are better at home, based on home care, it will reduce the cost of the care, plus it also results in a better outcome. That's one of the major factors driving this revolution.

One more area which is also driving is what I call interoperability and standardization of the ecosystems, and this is primarily regulated by Governments. What governments are now dictating whether it's hospitals, insurance companies, or any intermediaries in the space, is that they want to make sure that the same data of the patient is exchanged between these companies, and it's consistent. What that also means is that finally the patient will own their own medical records, and they can now go and take their medical records for better care, and better health possibilities. So, they have complete control of their data. But then the management of the data falls again with the insurance companies and hospitals, so that's driving more investments.

Another area that is personally very close to me is precision medicine. This is being demanded more from life sciences companies, and what I predict in the next 20-years based on what I've seen happening in our industry, is that in 20 years most of the medicines we take will be personalized for us, based on our genetic profile. Genome sequencing has become so cheap that what I believe is, you'll go in and with blood work, you'll do your genetic profile, and then you have it with you. And then whenever somebody is going to prescribe your medicine, they will look up your genetic profile, map something, and then start prescribing specific medicines that probably work for you. This is happening, and I'm very excited about this field as well.

So, to consolidate what I am saying; there are some changes where we have a deluge of data coming in from variables, IoT, sensors, and other things in this space. Second is the precision medicine. The third is the interoperability and standardization of the data. So that's driving a lot of investments in the healthcare sector.

This new data revolution has fascinating, especially coming from the IoT side of things, so the combination or harmonization of structured/ unstructured, it sounds like potentially data at motion at rest, the finite personal data with historical records. And bringing all this together to precision medicine, if you will, a decision point is pretty fascinating.

What other industries do you believe may be ripe for this same pattern of, new data revolution?

27:13

I think pretty much every industry needs to look at this data revolution and adapt to it, right? Because this is happening right now. And even if I want to say specific industries, I will go as far as saying that if they don't really work on this field and adapt to the new way of how the data is being processed in real-time, in streaming ways, they might not persist for long! It's a very bold statement to make, but I think companies across the industries need to look at what is happening and adapt to this new way of dealing with information.

As Momenta is an avid investor in the digital industrial space, I'm curious; what interesting startups are you seeing, especially in this DataOps space?

28:02

That's a very tricky question, but I'll try to answer very honestly. What I see, especially looking from an investor myself, and I'll talk about that as well, is that there's a lot of capital floating around both from venture capitalists, private equity, and high net worth individuals, and they're looking for the next cycle, because when I say next cycle, what I think is we are already at the peak of exits in a lot of these companies, especially data related companies or DataOps related companies. The surplus capital is chasing very few opportunities and looking for a new home. And in my view, if you're looking for investment in a B2B organization which is managing data or DataOps and other things, a couple of things you need to be looking at and which probably Momenta also does is, the size of the market, the size of the pain which you're addressing and then the product-market fit, and then, of course, the leadership.

But when you consider these factors, a lot of companies have already attracted valuations especially in the AI space, and then are struggling to meet those lofty valuations, which are there. And a clear example is C3.ai, which went IPO, and right now the valuation is one-third of what the IPO is. So, with prudent investors like you and Momenta, I would caution against the hype of AI. So, look at what is the market size, what is the market fit, what are the pains they are addressing, and whether the valuations meet those.

However coming back to your point, we do work with some of our partners, and I think they are at a good valuation, they are doing some incredible work; one is a Starburst, Cloudera went private because they were public and then they took them off the grid, so now they are private and are doing some incredible things on the cloud.

And as I mentioned previously, I am also an investor, and one of the things I invested in looking at in my field, and how I can mentor them, is a company based out of India called Earth Technologies, they're a SaaS company, (software as a service), and they are working on leveraging data to speed up construction and transform the construction industry. So that's one industry that has not yet been transformed, and there are incredible opportunities if we go there. As we speak there are some innovations happening in that industry where people are looking at prefabrication, just look at – I think it's Swift pods, where they're manufacturing whole bathrooms, and then they come and install the bathroom while you're constructing the home. So, it's completely prefabricated.

And the way the revolution is happening in that across the space, in terms of design, how they're building that is very heartening to see. We might probably see houses being built in days or weeks instead of months.

Those are great examples, and I remember in our pre-conversations you mentioned a source of pride, in that where India has largely been an IT services provider, a very valued one in most cases. You're seeing a move of the economy to more IT products.  Interestingly enough you guys represent an interesting example of that. Maybe you can say a few more words putting your India economic development hat on.

31:37

Yeah, and thanks for that. One of the things I'm incredibly passionate about in India is, the fact that when I moved back about eight years ago, I see this drive-in person which I call New India, which is, they want to get out of poverty, and this is across the board. People who are doing menial jobs, or people who are Uber drivers, people who want to make a difference in their life. And what they're seeing is the ticket for them to get out of poverty is education.

I love companies that are investing in education tech in India, where there is an online mechanism for teaching the kids. One of the incredible things that have happened in India in the last eight or nine years is this reach of technology. So, if you go to villages in India, they have access to what we call GEO evolution, there's a huge Indian company called Geo which has launched one of the cheapest rates of mobile connectivity in the world. We literally pay cents for each gigabyte. That's how cheap it has become. What that has done is it has exposed large amounts of the Indian population to the outside world and connected them, and now they want to take up the next step. What I see is the culture is shifting in India. Initially, we projected a services hub, there is labor arbitrage, and you would get services incredibly cheap in India.

What that is translating now is more into product innovation, so there is a lot of innovation happening on the ground, products are built at scale for Indian audiences and India, and then these products might come outside India and reach across the globe as they mature.

Thanks for stating that, but we are also incredibly proud that we are building innovative products out of India.

Thank you for that. It is interesting to see the metamorphosis happening there. Certainly, we're seeing it from our investment angle as well.

Finally, in closing, just a quick question, where do you find your personal inspiration?

33:44

Personally, I love teaching and mentoring people. When I teach, especially at the Business Schools, and when I conduct sessions and talks, what I do is I interact with these brilliant minds coming out of Indian schools, their aspirations, their take on the new-age problems; they bring some complexity, and that inspires me, such as, 'How do we tackle this?' They think completely differently than I would about a problem. So, a lot of my inspiration actually comes from teaching and interacting with these young minds, which I love.

The other thing, as I mentioned, is traveling within India. I was out of India for almost 15 years, and when I went back for me, it was a new India. And when I'm traveling, when I meet people and how they're trying to build new things, and a new culture, in fact, it's something called,  self-sufficient India or building things for India. That culture has been imbibed over the last seven or eight years.

What that means is that we will see more people getting out of poverty, and we will generate a lot of hyping in employment for people. And that is probably what drives me now, can I create more employment? Can I foster more people to take up entrepreneurship and leadership positions and drive change in how we see things?

You certainly have a good start in not only being a role model but just certainly demonstrating how that works, given the growth trajectory of a Modak.

Milind, thank you for spending this time with us today.

35:29

Oh, great, Ken, thanks for your time. It was a pleasure talking with you.

As well. So, this has been Milind Chitgupakar, co-founder and Chief Analytics Officer of Modak Analytics, and a leader in the new data revolution.

You can learn more about Modak Analytics @modak.com. Thank you for listening, and please join us next week for our next Momenta Digital Thread podcast.

Thank you, and have a great day.

[End]