Hugh Massie: Welcome everybody. I'm Hugh Massey, the Executive Chairman and Founder of DNA Behavior International. And today I'm delighted to be hosting another Behavioral Economics Today podcast. And in these podcasts, we're always endeavoring to address issues that are front and center of how psychology and economics intersect. In particular, we're very focused on how people make decisions and what drives those decisions. And today I'm delighted to have with me Snehal Marutka from Goldfried Consulting. And Snehal has had an exceptional career working as a data scientist. in some major organizational transformation projects centred on data and and and decision making. She's recently worked on projects at McKinsey, eTrade, Fiserf and you know is right there at the top of major corporations in addressing what is key to them functioning properly. and we're gonna talk a lot today about data and how it influences decision making. So welcome Sneha.
Snehal: Thank you, Yu. that was very kind introduction and it is my pleasure to be here with you. I have been following your work and very impressed with what you are doing. And a little bit about Me. So I'm Snehal Murudkar. As you introduced me, I'm a founder and CEO of the Gold Friend Consulting, and I serve enterprises into data and AI transformations. Over the past decades, I actually had a great opportunity to work on enterprise data, cloud, AI transformations for the companies ranging from Mackenzie, Morgan Stanley, E-Trade. by Surf, Ultium Packaging, which was a great manufacturing company. And throughout my career, I worked on many different initiatives, taking everything from supply chain logistics and finance to enterprise scale transformation into data, cloud, and recently in AI. And what all in all of these experiences and all of these industries, what is the most common factor? Is that we are trying to help organizations. To solve their complex problems, take these problems, make it simplify, and scale the solution for them to bring the measurable outcomes, measurable business outcomes. So today my effort is the same. So with my my company, GoFriend Consulting, I want I'm serving the enterprises into data and AI, enabling them, providing them as end-to-end solution from strategy and architecture to in through the implementation, governance, strong governance, and the adoption of the AI. which ultimately going to make them understand the emphasis on the AI at the enterprise scale.
Hugh Massie: Yeah, I think it's you know, it's fascinating because AI has become very quickly over the last three or three and a half years such a big topic. It's not in many ways completely new topic. There are people out there who always remind me, Well, I've been doing AI for twenty years. But but it's become a bigger topic. And I think that Maybe just to before we get into s you know some more specific issues, just describe what an enterprise project looks like. You know, taking the McKinsey example, not going into anything that's confidential, you where you you know you can't share, but you know, it's a I I think what I want you to bring out is it's a multi year engagement that's affecting every dimension of the business.
Snehal: That's true. throughout all these different organizations when work and especially at the massive firm like MA McKenzie, what we see is there are various different business units which are directly or indirectly connected to each other with the one single source of truth, which is your enterprise data. And They all have like whatever their definitions for the customer adaptability or customer retention or the business performance, they all have their own understanding of the business data. So at the enterprise scale, what is important is that not just collecting and one thing is that all of these are organizations they collect and they analyze and then they report on a massive Amount of data. There are a lot of processes that are massaging this data. There are a lot of processes that are analyzing the data. And there are a lot of people and business units that are reporting on this data. What is one thing that I've seen is most important thing is the communication strategy between the people. to understand this data, to have a trust in this data, having a strong governance because there are many people are connected with this data, having a shared business understanding between these units, business units that are working day to day together, whether it is operations, whether it is sales, HR, or technology or finance. So these all of the units when the when the bigger organization comes into picture, they are working together. Even though they're working on their priorities, they are connected together with the enterprise data.
Hugh Massie: Yeah, and I think it's you know, if you took the McKinsey example, and I'm be maybe being very simple here, but if you've got the the London office working on a banking and finance project, which they undoubtedly would do over there, with London being a big business centre and then and then the New York office, you want them singing off the same hymn sheets about data and banking and finance data, how it's interpreted and what information is provided to the end client and it would be very easy for that not to happen and therefore get incongruent results.
Snehal: which you said very dodge spot on because this is the same thing that we did and so in the what I have learned and my in my experience that the project, the initiative, the massive initiative, what we did is bringing all of this so there was a fragmented data all over different business units or different regions to s to that per se. And it was very imply important for the compliance reason that we bring our ERP system together, which was a massive finance cloud transformation. And at the same time. you touch my my finance cloud transformation you have to make sure your enterprise data is aligned with your ERP systems so we had to bring together three people 350 to 400 plus internal plus external clients together plus our entire entire enterprise together to make this massive change to to bring our legacy systems to the newer systems and that was the whole transformation of the entire company. So you can understand now as you speaking of the talking of the languages of London or New York, it is very important that organizations Follow the same standards, and that's where the ERP systems come and your compliance so that your company stayed in compliance with what they are and what they present to as an organization. and that is actually you were you hit the right right point there that it is very important to have that congruency throughout the organization.
Hugh Massie: Yeah, and it's a massive investment for the organization. And it's I think very easy for large organizations to get off track with data and and people just to, you know, accumulate loads of it and and file it somewhere and then it shows up in some way and someone goes and taps it for some purpose and you know, potentially the wrong decision gets made if that data is wrong.
Snehal: is. It is.
Hugh Massie: And I think that really brings me to the really big question that we probably should sort out up front in you know in our conversation now is with ai systems, large language models can go and read anything that they're allowed to to go and read, right? If you set them free, they will read everything. and try and make its own head or tail of what's been what's there and so I think where we're really getting to is the success of implementing AI and trusting AI comes down to data.
Snehal: Yes, you That is so important point about current technology. AI is extremely powerful. Google and Anthropic and OpenAI, they are coming with the latest models, which as you said are very powerful. They can read everything and anything. and that's why it is very important what is what your underlying data is, because the powerful data you feed, the powerful your business outcome is going to be. If you feed garbage in, it's gonna be a garbage out. So that is very important concept now because you need to have the trusted data, you need to have a strong governance around this data, and you need to have a shared business understanding between the different business units that is using this data. Because if you take an example of customer retention or you take an example of off. Operational efficiency. So sales or say technology or finance, they might have a different understanding of how their priorities are or how their applications are working with this data. And AI, you once you start using the AI, you need to make sure that your underlying data is as powerful as what you are projecting that it it is going to give you the output of of what your goals are. So that is very important thing that people are facing now that due to this fragmented data that has been created throughout over the years within these organizations, it is becoming a little difficult when when I talk to organizations, say for example, if it is a banking industry, they have a various different business unit for their checking account, for their saving, their loans, or their instruments. So these different business units have this data, but then Having the shared understanding to create their banking products, it is difficult because this data is very fragmented. There is no shared understanding between any of these business units. So AI needs this power from within of your enterprise data or to the matter of the data that you're feeding to the AI.
Hugh Massie: Yeah, and I think that, you know, if you break it down, the data needs to be accurate. And you use that term single source of truth. So, you know, if it's important that you know where somebody works and what job and tasks they do, you need to have that information properly recorded because it could be in the same organization, which I've seen, that's there's three different records of that.
Snehal: Exactly.
Hugh Massie: what's one which one is right, right? I mean that's a very basic you know, a basic example. And I think that's true in, you know, for a lot of banks. You know, one that we've we've been working with, they know the name of the person and how much money they've got in deposit accounts, but the other information about them is spread all over the place. And depending on what Salesforce accounts and other data sources that you go to to get
Snehal: That that is correct, yeah.
Hugh Massie: get the information, it's it's pretty messy stuff. and then is it actually the right data that you need? So it could what's there could be accurate, but is actually the right data is you know is another part. And then how accessible is the data? Because I think that people in organizations tend to put up walls. So if and everyone knows that data is important, but people have got different definitions of privacy or what their kingdom is, versus somebody else. And that all becomes a problem that that's got to be broken down. And someone's got to break that down.
Snehal: That is that is yeah, that is so true. And policies in different regions are different. They're compliance like GDPR in the Europe and we have our SAR compliance, various FINDRA rules. So they all all together it is very important. And in my experience I'll give an example, how we capture the the customers or vendors and and having their addresses and having their information stored. We had in one of the transformation that I did It is in CRM system it was captured differently. In the in the procurement department it is captured differently. It is stored in the enterprise data differently. And and that's where like when we were going through the transformation, we have to come together to consolidate all of this and to make sure that we are talking the same thing throughout so that we understand our customer well. And you are so right about that. And that's where it is very important to have the shared business understanding.
Hugh Massie: Yeah, and I think someone's gotta someone at the top's gotta set the rules for that or at least set the agenda. You know, and I always look at all these things maybe very simplistically, but it's got to come from the CEO to say that data's important, that we're gonna have a single source of truth. and it's an investment, right? you know, it's a key capital allocation to make sure it happens and sometimes some heads might be need to be banged together to realize that this data is the banks or the organizations, not a division heads, being brutal about it because I think that there's the next question is, you know, we if we establish that AI really cannot function without data and in a way AI is just technology itself. It's the programming of the AI that's important. That's that's humans as well. We'll get to that part maybe a little way along, but but we've got to have the right data. But then where does the data sit? And I think the the what really brought me to you and gotten me interested at an even deeper level than this was you know hearing about enterprise data warehouses. And I think for some people that might be considered not as relevant, but you know, with AI, but I think it's even more relevant.
Snehal: It it is. It is a single source of truth to your organization. Because what it does is like e the in the the enterprise data actually connects the all all your organiz all the entire organization together. There are so your all the different business units are directly or indirectly connected with your enterprise data because they are integrated with your it could be your ERP system, your CRM system, or your operations or your technology, but all of these different functions of your business. Are connected together with the enterprise data. The enterprise data warehouse basically is a single source of growth, which is in my experience where I have implemented this transformation. when we went through finance cloud transformation and the data transformation, the enterprise data was at the core. And what important thing it was that not just the technology that we wanted to we needed to invest we in we needed to invest in the communication strategy with the people that we work with. So And that that's why this enterprise data warehouse, when you're talking about this single source of truth with all of these people in your organization as well as external people that are connecting to this data, it becomes very important that your enterprise data is solid and it is a trusted and it is governed to talk the same language with within or outside of the company.
Hugh Massie: and the warehouse, like a w like a physical warehouse, you know, even though the enterprise data warehouse is going to be, you know, in the cloud, somewhere, it needs to be managed. and, you know, you use the word governance, it needs governance and it needs, you know, day to day management and processes to ensure that it's kept up to date and accurate. It doesn't really matter how big the business is. You know, I think that's something that's important too.
Snehal: It is and to tell you the truth in the projects that I have worked and the initiatives that we took upon, it we have spent a massive amount of time in governing the data because that is the most important factor for your compliance and for how people use this data within your organization or outside of your organization. So your RBAC or your role-based access or your attribute-based, we literally actually went to the level of the attribute level of how this data is going to be used within the organization. So governance is one of the most important factor and And it is connected with All of the employees within your organization, how each person is touching this data and how they are utilizing. So it is very important that the organization invest into the humans who are using this data. Technology, just like technology, we the people are the most important factor along with this technology. And that's where all of these factors come into picture. Your trusted data, the meaning of that data. The governance of that data, and as you said, we did actually spend a lot of time in the governing the data. That was the biggest factor when we went to the transformation.
Hugh Massie: And I think a you know look, a key reason to have centralized data in the enterprise data warehouse, get clean data, is to is ultimately to improve decision making. But just having good data I don't think just means that the outcome of that is there's going to be good decision making.
Snehal: And that is that is so true. like the leadership team the exec team they need to understand that just like how we invest into the technology how we invest into the understanding what business outcomes we are going to get they need to invest into the so we try to get around the metrics and we try to get around what what our goals are but to get the excellent business outcome it is where important that we invest into people who are understanding this data and the leadership has to have that communication strategy as a big umbrella for between between just the technology and and what your goals are.
Hugh Massie: Yeah, I think we're talking about in here, you know, the meaning of the data, you know, and how is it in how is it interpreted? And different humans can interpret the same data differently. it's a topic that I call and have written about called behavioral variability. so you know, even if we have the same data and and there's ten people sitting around the table all looking at it. And it's accurate and we know it's accurate, which helps. But the individual sitting around the table could still come up with different decisions about that data.
Snehal: I and that is true. That is very true.
Hugh Massie: de depending on their lens, and that's going to be some of that's systemic, you know, biases in terms of our behavioral biases, but also there's random factors, you know, in terms of how it how the data is laid out. and I think that gets to what you've you've been, you know, saying to me at varying times about the meaning of the data and how's it laid out so that it's interpreted in a particular way.
Snehal: And and that is that is really key key to that. Because, say, for example, we take the product performance. When you're talking about product performance, or you're talking about any metrics related to that, the sales have one definition, or the operations have their own definitions, technology comes with their own definitions, and then finance comes with their own definitions. And this was a true problem actually I faced when I went through various different organizations talking about whether I was in Morgan. Stanley or I I was at ETR or Pfizer or Mackenzie and Company or the manufacturing company. It it is one thing that that is That always has been in within the organization, that various different business units, they need they come up with their because they're because of their priorities. And what it is is that the exec level or the leadership level, it is that that's where it needs to start. That's where it needs to propagate throughout your organization to create and invest into the people to understand the business. meaning of what in what your data is because when the priorities are created and the metrics are laid down the goals and everything is laid down but the meaning is never propagated throughout from the leadership to the to the people What happens is like everybody creates their priorities, everybody works on their own initiatives, but what remains apart from all of this is having a consensus on what they are achieving, having a consensus or on the goals they are achieving and the data that they are understanding underneath this. So say for example there is a major initiative going throughout the organization, then it needs to propagate through from the leadership to the to the to all the levels of your organization to understand the shared meaning of business. And they should know what the product is all about.
Hugh Massie: And if you leave it to people's imaginations too much, they'll interpret it a particular way. and you know, it doesn't take very many layers of communication down through an organization for there to be ultimately complete chaos. and I think what we're talking about here for the audience is a concept of semantic debt. which is you know, the build up in a way of misinterpretations of data, you know, misunderstandings that then cascade their way through the business. So the data that the organization started with might have been okay, right? Might have been clean. But in a way that can get or that good work can get destroyed if you don't put in place you know these interpretation layers translation layers, how it's laid out, you know, the way it f you know the information flows, you don't get that right, you end up with this semantic debt problem, which is equivalent, but might be worse actually than the technology debt issue that organizations have with lots of legacy debt, you know, which is a cost and a blockage, the wrong data is going to lead to a lack of trust in decision making in the end.
Snehal: And that's what is the key actually. That's what is like what is blocking many different initiatives around the AI because organizations solving this problem that investing we invested in a lot of technology and throughout my throughout my career that I worked through supply chain, fintech, healthcare or manufacturing, what I see is like we have collected all the massive data. We have invested investment in done bigger investments in the technology, the latest cloud or AI. But what we have not invested in is having The priority on the semantic debt. The technology debt is the semantic debt is equally important as a technology debt. And you're spot on about that because we invest in everything in technology, but from if I I would be happy if we initially from the beginning, from the decades, we invested in the semantic debt.
Hugh Massie: I I'll be a little more controversial than you, Snehal, and you can tell me if I'm going too far. But I think the semantic debt is a more costly problem for the organization of technology debt because the technology debt can be quantified and steps can be taken to solve a short term problem with that. But with the semantic debt,
Snehal: Yes.
Hugh Massie: You may never know that there's the problem because the because data is invisible. human, you know, human behavior that goes with that is invisible and therefore decisions are made inadvertently on completely the wrong data. and the outcome of that is significant. You c it's hard to put a price on it. And I think that's where, you know, while Leaders like to think ego wise that they're great decision makers, and some are. But if there aren't with the wrong data, then this is a catastrophe and I think that's the semantic debt risk and we need to invest in it.
Snehal: And I agree with you, because that is the most important. That's why I'm saying, like from decades, I wish we invested more into equally well into the semantic than the tech I mean we invested into technology debt and that's why we moved on from legacy systems to cloud and now to cloud through ai so we are we are progressing in technology but I am but what we see is more mainly the semantic debt that is not that is getting complicated and
Hugh Massie: Yeah, and I think it's for some organizations, you know, it's it's well, we need to get the technology stack right before we can deal with the data and decision making and all the other things. And that I think that might have been true once upon a time, but I think nowadays, with data being so important, if not, you know, it's the foundation of AI, and AI, in a sense, is technology, both have got to be invested in equally. you know, and brought along inside the organization with an equal commitment and not one before the other. And I think so often when I turn up at organizations, it's like, well, we just need to deal with our CRM and the tech, you know, and our tech stack structure. but okay, fair enough. But you also need to deal with the data issue, because otherwise you're just gonna in the meantime going to keep on making wrong decisions.
Snehal: It is true, yeah. Yeah, I and that's what is very important because as I said, like I wish we invested into semantic debt from the beginning than just the technology debt because there has been a lot of investment done in the technology and which is fair because we need to move on from legacy systems to cloud and from cloud with cloud now the AI. but They are going to be powerful only if your data is powerful.
Hugh Massie: And I think that the the other point with semantic debt is, you know, i it's one thing for the raw data, if you want to call it that, to be clean, but the the issue, you know, around semantic debt is is the understanding, is the documentation of how data should be interpreted. And I think it's good in my mind, it's good for the listeners to this to understand that the semantic you know, the translation of the data and the policies and procedures in a way is data itself. and so it's data on data there and so it's a vicious circle that that's got to be put in order to
Snehal: And and and you you s said it very right because the trading company that I work it This is a mo one of the very important factors for us that whatever we do, we have to document every single thing so that your engineers, your leaders, or your managers understand the same language that that we are putting into the technology. So the documentation is one of the another key factor that that
Hugh Massie: Yeah.
Snehal: I always practiced in my career. so even the recent example that I worked on one of the AI projects for the healthcare industry, I began with defining what we are trying to do. I began with the proper documentation to to explain what is the factor that we are looking here in the project and what is our business outcome. So That is one of the one of the important things I feel that every organization needs to practice. That have a documentation that will give you a clear understanding of what is the product and what is the underlying definition of the goals that you are achieving with that product. So so that a so that every level, whether it is whether it is your C level, whether it is your VPs or therapists, or your engineering team or your finance team understand what what is happening with that particular product. So I mean and and that is one of the important thing that we practice throughout this throughout in my career at least, my team always followed having a documentation, having a documentation around your engineering initiatives because that is important thing. otherwise what happens is when you do not have any documented documented products that it the chaos gets pretty fast propagated throughout your different business unit to understand what is going on and they they start creating their own meanings out of it.
Hugh Massie: Right. And that's then when you've really got an out of control organization without without realizing it. And I think that's what we're getting to here and that's why the enterprise data warehouse is really the control point. If you're a CEO and the board of a company responsible for gu governance, you've got to have that. but but maybe if we get on to
Snehal: Yeah. Mm-hmm. Yeah.
Hugh Massie: New topic, and I think this is again a hidden part of all of this, and you've alluded to it a few times around communication, human beings. But what's missing from a lot of the data sets that an organization has?
Snehal: Mm-hmm. that's a very good question. and I I haven't no yeah no actually I I have been following your work for a few years and and what I learned from
Hugh Massie: I want you to do my work for me.
Snehal: your work plus what my experience over the decade is, human factor is really important. the technology is important, the business is important, and the human sits right on the top of that is very important because I feel there needs to be a big investment done in understanding people. Whether it is an employee understanding, whether it is a client understanding or whether it is a technology understanding Understanding. It is very important that we understand the humans. When they interact, the very good example is we were doing this massive transformation, and at that time we had to communicate with 350 plus customers, and it was internal and external customers. And at that time, they had their own priorities, but what we were touching is the entire enterprise data which was going. To affect all of these customers, and they they were not so customers when I talk customers, these are the internal teams who are going to work on the changes that we were doing through this transformation, plus the external customers who are gonna get affected with this and this enterprise transformations that we were doing. And at that time, it was very hard to bring everybody on board to emphasize that we are making this change. It is very important that you you have your deadlines, but you need to drop those and you need to focus on changing your systems to comply with the enterprise data changes that we are doing. And it was very difficult to to have them on board at the end. Niche at the beginning point of it. So I literally conducted health checks, call health check calls every week for this, which is now I'm talking about the communication strategy. Every week to health check on each of these customers. There were more than 300 plus consumers attending these calls to get on boarded. on what we are doing throughout the organization because it was a massive initiative and they needed to understand. So we literally had every week health check calls and during the production deployment we had everybody on board to test their systems. And then we conducted the testing every single week to make sure that and we we had a heat map to explain that are we doing this and who is falling behind. So to to to report them off of the heat map to let them know where you stand on this heat map. How important it's for you to make changes because once the whole system changes, whatever you're doing is not gonna work. So you need to make the changes as per what the whole organization is changing for. So and to that is to the point that it is the communication strategy and making these people to try while transforming the technology, transforming these people, making them understanding what is happening, these are the key factors and where the actually that's Where the employee understanding or the client understanding, all of these needs to be a big part of your initiatives, and that's where I I really enjoyed going through your work, and it it that's what it fits where I would say that layer would really work when people understand, and at the beginning, when any initiative starts, they create the whole plan and a strategy around. their technology plus the people.
Hugh Massie: Right, in a way there needs to be customized communication for each person, which comes from understanding their DNA style, which is what DNA behavior does. And then also by knowing that, you're also going to know their motivations, what appeals to each person. and then some people are creative and they can figure things out for themselves or have their way. And then it's others are extremely anchored and need a process or are cautious. or it's going to they think it's going to impact how they deal with a certain customer and and you need to be able to manage all of that to you know get everybody onto the same page so the communication is important but I think there's also an element here as well as you know when you're dealing with massive customer that bases and client bases is understanding the behavior of all of the customers so that you know and this is an area I'm
Snehal: Mm-hmm. Yeah.
Hugh Massie: passionate about it's one thing for an organization to have the financial transactional demographic data. That gets you so far. But what about the human? And I think Sneehal, you know, the example I often use in in my conversations is, you know, there can be two families living in the same cul-de-sac.
Snehal: Yeah.
Hugh Massie: They're they're in their early 40s, they've got a million dollars, they've got a home home paid off, and two kids. But banks and credit card companies and and and whoever else out there treat them all the same. But those people aren't the same. And how do the organizations recognize them as being different? And appeal to them, you know, in terms of communication style, maybe what their life preferences are. because you know, so one one of those couples might be very social, the other and like to do a lot of social things in their community and then the other couple and family want to save all of their money and be more quiet for a lot of the year and estudious, but then go and spend it all on life experiences. That's very different people and and different reasons for saving money or not saving money. but I think that the opportunity Is there now if you've got an enterprise data warehouse to humanize all of the data?
Snehal: Yeah, and that's where I can give you an example where the example you give is that the two different people having the same situation of financials but having a different different mindset, which is which is actually a key in in anything that you do and whether that is your taking a banking product or w whether that is a trading Product or or day-to-day operations of any human being, it is very important to understand the mindset. So, for example, like if you are if your product is getting launched, it is very important to understand a different control group of your clients and your different personas, and that's where, like, whenever we have launched our products, it was very important. important initially to understand our client mindset that what persona are we hitting and if it is a group of female or if it is a group of male of a certain age, how the youngsters are doing, whether how there are 50 plus communities are working with this product or so taking a loan products that are out there. what would the a person who is cre who is at the stage of starting his new family what is his needs versus a student who is going to a college what is what is the loan needs for the student so all of this understanding the persona and understanding the mindset of your client which is very important and in One of the digital banking products that we were when we were launching, we had done actually a lot of study on a client and a client personas. So to understand when this product is going to be out, how how it was a banking product about the notification about the fraud, fraud or a low balance, or giving them the notification about the new products of the banks that are coming out. So we were launching that and at that time we basically tried to understand the different persona of the people. How would they react to this particular product?
Hugh Massie: you're on top of the idea of understanding the mindset of the client. And I think that till now a lot of organizations have been limited in doing that to demo gr to basing that off demographic and financial information and making it an inference, that that is what tells you about the person. Whereas I think now with AI and what we've developed For example, at DNA Behavior with our digital scan product, for example, you can get to the psychometrics depth of the person, you know, to understand more about their inherent values, communication style, motivations, that then will tell you more about the person than just what's in their bank account. and I think that's where data can really become a huge asset for an organization. Particularly if you believe that, you know, owning the client experience is what's going to drive an organization or customer experience, what's going to drive the organization forward. Data and behavioral data are are assets. And really they should be booked on the balance sheet. If you're an accountant like I am, or was, should be booked on the balance sheet as an asset. it's the value of the business.
Snehal: Yeah. It is it is it is a great value of the business and it's the most important factor actually because anytime that we work and in my experience basically when I have worked through, we have created the whole a line graph to understand What what are we producing and how each business unit is going to interact with that data? And creating a different persona of each different department. and it is now when you're talking about the business unit, so it's a by the persona by the business unit. How would the operations work with this? How would the optimized department work with this, or how will the procurement department work would work with this data? So we had created a literally a your personal, the people interacting with the data.
Hugh Massie: And in a way, I think you're you're opening up the king hit point here that we've been unpacking and sort of I think in a way Snee How could be a concluding point is just knowing demographic information and transactional information and reporting that that's historical. But if you can know
Snehal: Mm-hmm. Mm-hmm.
Hugh Massie: more about the inner motivations of people.
Snehal: Mm.
Hugh Massie: And what they'll do it becomes predictive.
Snehal: It is, yeah, and and that is another factor of the s for the healthcare project that they understanding the people mindset is very important to do the predictive analysis of how would they react to a certain healthcare product. So not revealing the details, but it is it is important for whenever the healthcare product comes out. It is you you have to create a different age groups, different different health conditions and and the personas around that to understand how would they react with that particular product healthcare product.
Hugh Massie: know, so as we wrap up today, and I wanna leave the audience with one more thought, but today we've you know, we've talked a lot about the importance of data to the future of AI and the successful future of AI, which I think we all want AI to be successful and trustworthy. And I think trust is is the key thing. Enterprise data warehouses, applying or attributing the right meaning to data and documenting it, governance, humanizing data. You know, we've got all the key elements there. you know, that semantic debt is as bigger issue as or bigger issue than technical debt or technological debt. But one thing that I think should not be lost on anybody, particularly with AI, is it's the people behind the design of the AI is also important to understand because that's a decision.
Snehal: Yes, yeah.
Hugh Massie: And and and how AI is designed, if that comes from one person, in a way they're imputing their view and perspectives for everybody else to follow in terms of how decisions might get followed, how a product gets taken up, or what something actually means. You you could be educating the world, right? through one lens. And I think that we w sh nothing should ever be forgotten that AI is programmed by humans. even if it is AI programming AI, there is a human there somewhere and that makes it behavioural.
Snehal: It is true. It is true. and that's what I would say like Instead of isolating AI in into any isolated use cases, it needs to be together with considering human factor and what is the best asset for any organization is having the trusted data, having a strong governance around it, having the shared business meaning, and and invest the same into cementing that along with the technology there. because the powerful your data is going to be the powerful AI can create the business outcomes for you. And that's the key. And and that this is going to help the leadership into the decision making, the right decision making for for their organization. This is going to be very powerful for their decision-making strategy and also to any initiatives that needs to start be started. This factor needs to be embedded at the beginning of any initiative.
Hugh Massie: Yeah, and it's with the right with the right investment. So I think AI can be very powerful and successful if all of these steps are followed and there's and but if one just treats AI as magic and that AI is going to get it right, that's also an even bigger problem. that that mindset's bigger bigger problem than the semantic debt problem, I think.
Snehal: Yeah, yeah, yeah. And And one of the examples I'll give you, I was doing my postgraduation program with the UT Austin, and we did the seven real world projects. In that the first most important factor is like curating your data, completely creating a very solid understanding of what your data is, and and then understanding your persona and then creating your AI product out of that. So as all throughout this our discussion as we spoke, it is very important that you have a trusted data, you have strong governance, you have shared business meaning, and your semantic debt is taken care of. It's gonna give you the best outcomes, best decision making for leadership. and it's and we are gonna thrive in this AI era where AI is so powerful that it can read everything and anything that you have in your organization. So it is very important that you have the powerful data to create a powerful AI products.
Hugh Massie: And I I think that's a great message, Snihal, to finish up on and I just want to thank you. You know, you are in my mind a world authority on data and organizational transformation and empowering AI in a way, you know, I think so far you're one of maybe the unsung heroes because everyone talks about, you know, anthropic, but what's underneath anthropic, you know, and you're one of the powerhouses that are going to drive the world forward the right way. I think that it can clearly only come from data. So thank you for participating, you know, with me today in the conversation. I really appreciate it and I think that our audience is going to get a lot out of this in terms of what they need to do and be thinking about.
Snehal: Thank you. I mean it was my pleasure and truly have a great pleasure talking to you. And this is these are really current topics. These are very important topics that people are trying to understand. And I've been following your and very much impressed with what you are doing at the DNA behavior. So it was truly a pleasure to talk about the behavioral science plus like understanding like the human factor and the data with it. So it was really great discussion with you. I I truly enjoyed it.
Hugh Massie: So Sneehal, just just to wrap up, where should people find you if they want to find out more about you and your work, what's the best place for them to go?
Snehal: so I have my website goldfriendconsult dot com. And my email address which I can provide you it is SNPall my first name sne-h a l dot mururkar the last name M-U-R-U-D-K-A-R at goldfriendconsult.com. so they can reach out to me there and I I'll definitely answer or help them in their data and AI needs for their enterprise level initiatives.
Hugh Massie: Fantastic. There'll be plenty of them.
Snehal: Yeah, great.
Snehal: Thank you, Yu. that was very kind introduction and it is my pleasure to be here with you. I have been following your work and very impressed with what you are doing. And a little bit about Me. So I'm Snehal Murudkar. As you introduced me, I'm a founder and CEO of the Gold Friend Consulting, and I serve enterprises into data and AI transformations. Over the past decades, I actually had a great opportunity to work on enterprise data, cloud, AI transformations for the companies ranging from Mackenzie, Morgan Stanley, E-Trade. by Surf, Ultium Packaging, which was a great manufacturing company. And throughout my career, I worked on many different initiatives, taking everything from supply chain logistics and finance to enterprise scale transformation into data, cloud, and recently in AI. And what all in all of these experiences and all of these industries, what is the most common factor? Is that we are trying to help organizations. To solve their complex problems, take these problems, make it simplify, and scale the solution for them to bring the measurable outcomes, measurable business outcomes. So today my effort is the same. So with my my company, GoFriend Consulting, I want I'm serving the enterprises into data and AI, enabling them, providing them as end-to-end solution from strategy and architecture to in through the implementation, governance, strong governance, and the adoption of the AI. which ultimately going to make them understand the emphasis on the AI at the enterprise scale.
Hugh Massie: Yeah, I think it's you know, it's fascinating because AI has become very quickly over the last three or three and a half years such a big topic. It's not in many ways completely new topic. There are people out there who always remind me, Well, I've been doing AI for twenty years. But but it's become a bigger topic. And I think that Maybe just to before we get into s you know some more specific issues, just describe what an enterprise project looks like. You know, taking the McKinsey example, not going into anything that's confidential, you where you you know you can't share, but you know, it's a I I think what I want you to bring out is it's a multi year engagement that's affecting every dimension of the business.
Snehal: That's true. throughout all these different organizations when work and especially at the massive firm like MA McKenzie, what we see is there are various different business units which are directly or indirectly connected to each other with the one single source of truth, which is your enterprise data. And They all have like whatever their definitions for the customer adaptability or customer retention or the business performance, they all have their own understanding of the business data. So at the enterprise scale, what is important is that not just collecting and one thing is that all of these are organizations they collect and they analyze and then they report on a massive Amount of data. There are a lot of processes that are massaging this data. There are a lot of processes that are analyzing the data. And there are a lot of people and business units that are reporting on this data. What is one thing that I've seen is most important thing is the communication strategy between the people. to understand this data, to have a trust in this data, having a strong governance because there are many people are connected with this data, having a shared business understanding between these units, business units that are working day to day together, whether it is operations, whether it is sales, HR, or technology or finance. So these all of the units when the when the bigger organization comes into picture, they are working together. Even though they're working on their priorities, they are connected together with the enterprise data.
Hugh Massie: Yeah, and I think it's you know, if you took the McKinsey example, and I'm be maybe being very simple here, but if you've got the the London office working on a banking and finance project, which they undoubtedly would do over there, with London being a big business centre and then and then the New York office, you want them singing off the same hymn sheets about data and banking and finance data, how it's interpreted and what information is provided to the end client and it would be very easy for that not to happen and therefore get incongruent results.
Snehal: which you said very dodge spot on because this is the same thing that we did and so in the what I have learned and my in my experience that the project, the initiative, the massive initiative, what we did is bringing all of this so there was a fragmented data all over different business units or different regions to s to that per se. And it was very imply important for the compliance reason that we bring our ERP system together, which was a massive finance cloud transformation. And at the same time. you touch my my finance cloud transformation you have to make sure your enterprise data is aligned with your ERP systems so we had to bring together three people 350 to 400 plus internal plus external clients together plus our entire entire enterprise together to make this massive change to to bring our legacy systems to the newer systems and that was the whole transformation of the entire company. So you can understand now as you speaking of the talking of the languages of London or New York, it is very important that organizations Follow the same standards, and that's where the ERP systems come and your compliance so that your company stayed in compliance with what they are and what they present to as an organization. and that is actually you were you hit the right right point there that it is very important to have that congruency throughout the organization.
Hugh Massie: Yeah, and it's a massive investment for the organization. And it's I think very easy for large organizations to get off track with data and and people just to, you know, accumulate loads of it and and file it somewhere and then it shows up in some way and someone goes and taps it for some purpose and you know, potentially the wrong decision gets made if that data is wrong.
Snehal: is. It is.
Hugh Massie: And I think that really brings me to the really big question that we probably should sort out up front in you know in our conversation now is with ai systems, large language models can go and read anything that they're allowed to to go and read, right? If you set them free, they will read everything. and try and make its own head or tail of what's been what's there and so I think where we're really getting to is the success of implementing AI and trusting AI comes down to data.
Snehal: Yes, you That is so important point about current technology. AI is extremely powerful. Google and Anthropic and OpenAI, they are coming with the latest models, which as you said are very powerful. They can read everything and anything. and that's why it is very important what is what your underlying data is, because the powerful data you feed, the powerful your business outcome is going to be. If you feed garbage in, it's gonna be a garbage out. So that is very important concept now because you need to have the trusted data, you need to have a strong governance around this data, and you need to have a shared business understanding between the different business units that is using this data. Because if you take an example of customer retention or you take an example of off. Operational efficiency. So sales or say technology or finance, they might have a different understanding of how their priorities are or how their applications are working with this data. And AI, you once you start using the AI, you need to make sure that your underlying data is as powerful as what you are projecting that it it is going to give you the output of of what your goals are. So that is very important thing that people are facing now that due to this fragmented data that has been created throughout over the years within these organizations, it is becoming a little difficult when when I talk to organizations, say for example, if it is a banking industry, they have a various different business unit for their checking account, for their saving, their loans, or their instruments. So these different business units have this data, but then Having the shared understanding to create their banking products, it is difficult because this data is very fragmented. There is no shared understanding between any of these business units. So AI needs this power from within of your enterprise data or to the matter of the data that you're feeding to the AI.
Hugh Massie: Yeah, and I think that, you know, if you break it down, the data needs to be accurate. And you use that term single source of truth. So, you know, if it's important that you know where somebody works and what job and tasks they do, you need to have that information properly recorded because it could be in the same organization, which I've seen, that's there's three different records of that.
Snehal: Exactly.
Hugh Massie: what's one which one is right, right? I mean that's a very basic you know, a basic example. And I think that's true in, you know, for a lot of banks. You know, one that we've we've been working with, they know the name of the person and how much money they've got in deposit accounts, but the other information about them is spread all over the place. And depending on what Salesforce accounts and other data sources that you go to to get
Snehal: That that is correct, yeah.
Hugh Massie: get the information, it's it's pretty messy stuff. and then is it actually the right data that you need? So it could what's there could be accurate, but is actually the right data is you know is another part. And then how accessible is the data? Because I think that people in organizations tend to put up walls. So if and everyone knows that data is important, but people have got different definitions of privacy or what their kingdom is, versus somebody else. And that all becomes a problem that that's got to be broken down. And someone's got to break that down.
Snehal: That is that is yeah, that is so true. And policies in different regions are different. They're compliance like GDPR in the Europe and we have our SAR compliance, various FINDRA rules. So they all all together it is very important. And in my experience I'll give an example, how we capture the the customers or vendors and and having their addresses and having their information stored. We had in one of the transformation that I did It is in CRM system it was captured differently. In the in the procurement department it is captured differently. It is stored in the enterprise data differently. And and that's where like when we were going through the transformation, we have to come together to consolidate all of this and to make sure that we are talking the same thing throughout so that we understand our customer well. And you are so right about that. And that's where it is very important to have the shared business understanding.
Hugh Massie: Yeah, and I think someone's gotta someone at the top's gotta set the rules for that or at least set the agenda. You know, and I always look at all these things maybe very simplistically, but it's got to come from the CEO to say that data's important, that we're gonna have a single source of truth. and it's an investment, right? you know, it's a key capital allocation to make sure it happens and sometimes some heads might be need to be banged together to realize that this data is the banks or the organizations, not a division heads, being brutal about it because I think that there's the next question is, you know, we if we establish that AI really cannot function without data and in a way AI is just technology itself. It's the programming of the AI that's important. That's that's humans as well. We'll get to that part maybe a little way along, but but we've got to have the right data. But then where does the data sit? And I think the the what really brought me to you and gotten me interested at an even deeper level than this was you know hearing about enterprise data warehouses. And I think for some people that might be considered not as relevant, but you know, with AI, but I think it's even more relevant.
Snehal: It it is. It is a single source of truth to your organization. Because what it does is like e the in the the enterprise data actually connects the all all your organiz all the entire organization together. There are so your all the different business units are directly or indirectly connected with your enterprise data because they are integrated with your it could be your ERP system, your CRM system, or your operations or your technology, but all of these different functions of your business. Are connected together with the enterprise data. The enterprise data warehouse basically is a single source of growth, which is in my experience where I have implemented this transformation. when we went through finance cloud transformation and the data transformation, the enterprise data was at the core. And what important thing it was that not just the technology that we wanted to we needed to invest we in we needed to invest in the communication strategy with the people that we work with. So And that that's why this enterprise data warehouse, when you're talking about this single source of truth with all of these people in your organization as well as external people that are connecting to this data, it becomes very important that your enterprise data is solid and it is a trusted and it is governed to talk the same language with within or outside of the company.
Hugh Massie: and the warehouse, like a w like a physical warehouse, you know, even though the enterprise data warehouse is going to be, you know, in the cloud, somewhere, it needs to be managed. and, you know, you use the word governance, it needs governance and it needs, you know, day to day management and processes to ensure that it's kept up to date and accurate. It doesn't really matter how big the business is. You know, I think that's something that's important too.
Snehal: It is and to tell you the truth in the projects that I have worked and the initiatives that we took upon, it we have spent a massive amount of time in governing the data because that is the most important factor for your compliance and for how people use this data within your organization or outside of your organization. So your RBAC or your role-based access or your attribute-based, we literally actually went to the level of the attribute level of how this data is going to be used within the organization. So governance is one of the most important factor and And it is connected with All of the employees within your organization, how each person is touching this data and how they are utilizing. So it is very important that the organization invest into the humans who are using this data. Technology, just like technology, we the people are the most important factor along with this technology. And that's where all of these factors come into picture. Your trusted data, the meaning of that data. The governance of that data, and as you said, we did actually spend a lot of time in the governing the data. That was the biggest factor when we went to the transformation.
Hugh Massie: And I think a you know look, a key reason to have centralized data in the enterprise data warehouse, get clean data, is to is ultimately to improve decision making. But just having good data I don't think just means that the outcome of that is there's going to be good decision making.
Snehal: And that is that is so true. like the leadership team the exec team they need to understand that just like how we invest into the technology how we invest into the understanding what business outcomes we are going to get they need to invest into the so we try to get around the metrics and we try to get around what what our goals are but to get the excellent business outcome it is where important that we invest into people who are understanding this data and the leadership has to have that communication strategy as a big umbrella for between between just the technology and and what your goals are.
Hugh Massie: Yeah, I think we're talking about in here, you know, the meaning of the data, you know, and how is it in how is it interpreted? And different humans can interpret the same data differently. it's a topic that I call and have written about called behavioral variability. so you know, even if we have the same data and and there's ten people sitting around the table all looking at it. And it's accurate and we know it's accurate, which helps. But the individual sitting around the table could still come up with different decisions about that data.
Snehal: I and that is true. That is very true.
Hugh Massie: de depending on their lens, and that's going to be some of that's systemic, you know, biases in terms of our behavioral biases, but also there's random factors, you know, in terms of how it how the data is laid out. and I think that gets to what you've you've been, you know, saying to me at varying times about the meaning of the data and how's it laid out so that it's interpreted in a particular way.
Snehal: And and that is that is really key key to that. Because, say, for example, we take the product performance. When you're talking about product performance, or you're talking about any metrics related to that, the sales have one definition, or the operations have their own definitions, technology comes with their own definitions, and then finance comes with their own definitions. And this was a true problem actually I faced when I went through various different organizations talking about whether I was in Morgan. Stanley or I I was at ETR or Pfizer or Mackenzie and Company or the manufacturing company. It it is one thing that that is That always has been in within the organization, that various different business units, they need they come up with their because they're because of their priorities. And what it is is that the exec level or the leadership level, it is that that's where it needs to start. That's where it needs to propagate throughout your organization to create and invest into the people to understand the business. meaning of what in what your data is because when the priorities are created and the metrics are laid down the goals and everything is laid down but the meaning is never propagated throughout from the leadership to the to the people What happens is like everybody creates their priorities, everybody works on their own initiatives, but what remains apart from all of this is having a consensus on what they are achieving, having a consensus or on the goals they are achieving and the data that they are understanding underneath this. So say for example there is a major initiative going throughout the organization, then it needs to propagate through from the leadership to the to the to all the levels of your organization to understand the shared meaning of business. And they should know what the product is all about.
Hugh Massie: And if you leave it to people's imaginations too much, they'll interpret it a particular way. and you know, it doesn't take very many layers of communication down through an organization for there to be ultimately complete chaos. and I think what we're talking about here for the audience is a concept of semantic debt. which is you know, the build up in a way of misinterpretations of data, you know, misunderstandings that then cascade their way through the business. So the data that the organization started with might have been okay, right? Might have been clean. But in a way that can get or that good work can get destroyed if you don't put in place you know these interpretation layers translation layers, how it's laid out, you know, the way it f you know the information flows, you don't get that right, you end up with this semantic debt problem, which is equivalent, but might be worse actually than the technology debt issue that organizations have with lots of legacy debt, you know, which is a cost and a blockage, the wrong data is going to lead to a lack of trust in decision making in the end.
Snehal: And that's what is the key actually. That's what is like what is blocking many different initiatives around the AI because organizations solving this problem that investing we invested in a lot of technology and throughout my throughout my career that I worked through supply chain, fintech, healthcare or manufacturing, what I see is like we have collected all the massive data. We have invested investment in done bigger investments in the technology, the latest cloud or AI. But what we have not invested in is having The priority on the semantic debt. The technology debt is the semantic debt is equally important as a technology debt. And you're spot on about that because we invest in everything in technology, but from if I I would be happy if we initially from the beginning, from the decades, we invested in the semantic debt.
Hugh Massie: I I'll be a little more controversial than you, Snehal, and you can tell me if I'm going too far. But I think the semantic debt is a more costly problem for the organization of technology debt because the technology debt can be quantified and steps can be taken to solve a short term problem with that. But with the semantic debt,
Snehal: Yes.
Hugh Massie: You may never know that there's the problem because the because data is invisible. human, you know, human behavior that goes with that is invisible and therefore decisions are made inadvertently on completely the wrong data. and the outcome of that is significant. You c it's hard to put a price on it. And I think that's where, you know, while Leaders like to think ego wise that they're great decision makers, and some are. But if there aren't with the wrong data, then this is a catastrophe and I think that's the semantic debt risk and we need to invest in it.
Snehal: And I agree with you, because that is the most important. That's why I'm saying, like from decades, I wish we invested more into equally well into the semantic than the tech I mean we invested into technology debt and that's why we moved on from legacy systems to cloud and now to cloud through ai so we are we are progressing in technology but I am but what we see is more mainly the semantic debt that is not that is getting complicated and
Hugh Massie: Yeah, and I think it's for some organizations, you know, it's it's well, we need to get the technology stack right before we can deal with the data and decision making and all the other things. And that I think that might have been true once upon a time, but I think nowadays, with data being so important, if not, you know, it's the foundation of AI, and AI, in a sense, is technology, both have got to be invested in equally. you know, and brought along inside the organization with an equal commitment and not one before the other. And I think so often when I turn up at organizations, it's like, well, we just need to deal with our CRM and the tech, you know, and our tech stack structure. but okay, fair enough. But you also need to deal with the data issue, because otherwise you're just gonna in the meantime going to keep on making wrong decisions.
Snehal: It is true, yeah. Yeah, I and that's what is very important because as I said, like I wish we invested into semantic debt from the beginning than just the technology debt because there has been a lot of investment done in the technology and which is fair because we need to move on from legacy systems to cloud and from cloud with cloud now the AI. but They are going to be powerful only if your data is powerful.
Hugh Massie: And I think that the the other point with semantic debt is, you know, i it's one thing for the raw data, if you want to call it that, to be clean, but the the issue, you know, around semantic debt is is the understanding, is the documentation of how data should be interpreted. And I think it's good in my mind, it's good for the listeners to this to understand that the semantic you know, the translation of the data and the policies and procedures in a way is data itself. and so it's data on data there and so it's a vicious circle that that's got to be put in order to
Snehal: And and and you you s said it very right because the trading company that I work it This is a mo one of the very important factors for us that whatever we do, we have to document every single thing so that your engineers, your leaders, or your managers understand the same language that that we are putting into the technology. So the documentation is one of the another key factor that that
Hugh Massie: Yeah.
Snehal: I always practiced in my career. so even the recent example that I worked on one of the AI projects for the healthcare industry, I began with defining what we are trying to do. I began with the proper documentation to to explain what is the factor that we are looking here in the project and what is our business outcome. So That is one of the one of the important things I feel that every organization needs to practice. That have a documentation that will give you a clear understanding of what is the product and what is the underlying definition of the goals that you are achieving with that product. So so that a so that every level, whether it is whether it is your C level, whether it is your VPs or therapists, or your engineering team or your finance team understand what what is happening with that particular product. So I mean and and that is one of the important thing that we practice throughout this throughout in my career at least, my team always followed having a documentation, having a documentation around your engineering initiatives because that is important thing. otherwise what happens is when you do not have any documented documented products that it the chaos gets pretty fast propagated throughout your different business unit to understand what is going on and they they start creating their own meanings out of it.
Hugh Massie: Right. And that's then when you've really got an out of control organization without without realizing it. And I think that's what we're getting to here and that's why the enterprise data warehouse is really the control point. If you're a CEO and the board of a company responsible for gu governance, you've got to have that. but but maybe if we get on to
Snehal: Yeah. Mm-hmm. Yeah.
Hugh Massie: New topic, and I think this is again a hidden part of all of this, and you've alluded to it a few times around communication, human beings. But what's missing from a lot of the data sets that an organization has?
Snehal: Mm-hmm. that's a very good question. and I I haven't no yeah no actually I I have been following your work for a few years and and what I learned from
Hugh Massie: I want you to do my work for me.
Snehal: your work plus what my experience over the decade is, human factor is really important. the technology is important, the business is important, and the human sits right on the top of that is very important because I feel there needs to be a big investment done in understanding people. Whether it is an employee understanding, whether it is a client understanding or whether it is a technology understanding Understanding. It is very important that we understand the humans. When they interact, the very good example is we were doing this massive transformation, and at that time we had to communicate with 350 plus customers, and it was internal and external customers. And at that time, they had their own priorities, but what we were touching is the entire enterprise data which was going. To affect all of these customers, and they they were not so customers when I talk customers, these are the internal teams who are going to work on the changes that we were doing through this transformation, plus the external customers who are gonna get affected with this and this enterprise transformations that we were doing. And at that time, it was very hard to bring everybody on board to emphasize that we are making this change. It is very important that you you have your deadlines, but you need to drop those and you need to focus on changing your systems to comply with the enterprise data changes that we are doing. And it was very difficult to to have them on board at the end. Niche at the beginning point of it. So I literally conducted health checks, call health check calls every week for this, which is now I'm talking about the communication strategy. Every week to health check on each of these customers. There were more than 300 plus consumers attending these calls to get on boarded. on what we are doing throughout the organization because it was a massive initiative and they needed to understand. So we literally had every week health check calls and during the production deployment we had everybody on board to test their systems. And then we conducted the testing every single week to make sure that and we we had a heat map to explain that are we doing this and who is falling behind. So to to to report them off of the heat map to let them know where you stand on this heat map. How important it's for you to make changes because once the whole system changes, whatever you're doing is not gonna work. So you need to make the changes as per what the whole organization is changing for. So and to that is to the point that it is the communication strategy and making these people to try while transforming the technology, transforming these people, making them understanding what is happening, these are the key factors and where the actually that's Where the employee understanding or the client understanding, all of these needs to be a big part of your initiatives, and that's where I I really enjoyed going through your work, and it it that's what it fits where I would say that layer would really work when people understand, and at the beginning, when any initiative starts, they create the whole plan and a strategy around. their technology plus the people.
Hugh Massie: Right, in a way there needs to be customized communication for each person, which comes from understanding their DNA style, which is what DNA behavior does. And then also by knowing that, you're also going to know their motivations, what appeals to each person. and then some people are creative and they can figure things out for themselves or have their way. And then it's others are extremely anchored and need a process or are cautious. or it's going to they think it's going to impact how they deal with a certain customer and and you need to be able to manage all of that to you know get everybody onto the same page so the communication is important but I think there's also an element here as well as you know when you're dealing with massive customer that bases and client bases is understanding the behavior of all of the customers so that you know and this is an area I'm
Snehal: Mm-hmm. Yeah.
Hugh Massie: passionate about it's one thing for an organization to have the financial transactional demographic data. That gets you so far. But what about the human? And I think Sneehal, you know, the example I often use in in my conversations is, you know, there can be two families living in the same cul-de-sac.
Snehal: Yeah.
Hugh Massie: They're they're in their early 40s, they've got a million dollars, they've got a home home paid off, and two kids. But banks and credit card companies and and and whoever else out there treat them all the same. But those people aren't the same. And how do the organizations recognize them as being different? And appeal to them, you know, in terms of communication style, maybe what their life preferences are. because you know, so one one of those couples might be very social, the other and like to do a lot of social things in their community and then the other couple and family want to save all of their money and be more quiet for a lot of the year and estudious, but then go and spend it all on life experiences. That's very different people and and different reasons for saving money or not saving money. but I think that the opportunity Is there now if you've got an enterprise data warehouse to humanize all of the data?
Snehal: Yeah, and that's where I can give you an example where the example you give is that the two different people having the same situation of financials but having a different different mindset, which is which is actually a key in in anything that you do and whether that is your taking a banking product or w whether that is a trading Product or or day-to-day operations of any human being, it is very important to understand the mindset. So, for example, like if you are if your product is getting launched, it is very important to understand a different control group of your clients and your different personas, and that's where, like, whenever we have launched our products, it was very important. important initially to understand our client mindset that what persona are we hitting and if it is a group of female or if it is a group of male of a certain age, how the youngsters are doing, whether how there are 50 plus communities are working with this product or so taking a loan products that are out there. what would the a person who is cre who is at the stage of starting his new family what is his needs versus a student who is going to a college what is what is the loan needs for the student so all of this understanding the persona and understanding the mindset of your client which is very important and in One of the digital banking products that we were when we were launching, we had done actually a lot of study on a client and a client personas. So to understand when this product is going to be out, how how it was a banking product about the notification about the fraud, fraud or a low balance, or giving them the notification about the new products of the banks that are coming out. So we were launching that and at that time we basically tried to understand the different persona of the people. How would they react to this particular product?
Hugh Massie: you're on top of the idea of understanding the mindset of the client. And I think that till now a lot of organizations have been limited in doing that to demo gr to basing that off demographic and financial information and making it an inference, that that is what tells you about the person. Whereas I think now with AI and what we've developed For example, at DNA Behavior with our digital scan product, for example, you can get to the psychometrics depth of the person, you know, to understand more about their inherent values, communication style, motivations, that then will tell you more about the person than just what's in their bank account. and I think that's where data can really become a huge asset for an organization. Particularly if you believe that, you know, owning the client experience is what's going to drive an organization or customer experience, what's going to drive the organization forward. Data and behavioral data are are assets. And really they should be booked on the balance sheet. If you're an accountant like I am, or was, should be booked on the balance sheet as an asset. it's the value of the business.
Snehal: Yeah. It is it is it is a great value of the business and it's the most important factor actually because anytime that we work and in my experience basically when I have worked through, we have created the whole a line graph to understand What what are we producing and how each business unit is going to interact with that data? And creating a different persona of each different department. and it is now when you're talking about the business unit, so it's a by the persona by the business unit. How would the operations work with this? How would the optimized department work with this, or how will the procurement department work would work with this data? So we had created a literally a your personal, the people interacting with the data.
Hugh Massie: And in a way, I think you're you're opening up the king hit point here that we've been unpacking and sort of I think in a way Snee How could be a concluding point is just knowing demographic information and transactional information and reporting that that's historical. But if you can know
Snehal: Mm-hmm. Mm-hmm.
Hugh Massie: more about the inner motivations of people.
Snehal: Mm.
Hugh Massie: And what they'll do it becomes predictive.
Snehal: It is, yeah, and and that is another factor of the s for the healthcare project that they understanding the people mindset is very important to do the predictive analysis of how would they react to a certain healthcare product. So not revealing the details, but it is it is important for whenever the healthcare product comes out. It is you you have to create a different age groups, different different health conditions and and the personas around that to understand how would they react with that particular product healthcare product.
Hugh Massie: know, so as we wrap up today, and I wanna leave the audience with one more thought, but today we've you know, we've talked a lot about the importance of data to the future of AI and the successful future of AI, which I think we all want AI to be successful and trustworthy. And I think trust is is the key thing. Enterprise data warehouses, applying or attributing the right meaning to data and documenting it, governance, humanizing data. You know, we've got all the key elements there. you know, that semantic debt is as bigger issue as or bigger issue than technical debt or technological debt. But one thing that I think should not be lost on anybody, particularly with AI, is it's the people behind the design of the AI is also important to understand because that's a decision.
Snehal: Yes, yeah.
Hugh Massie: And and and how AI is designed, if that comes from one person, in a way they're imputing their view and perspectives for everybody else to follow in terms of how decisions might get followed, how a product gets taken up, or what something actually means. You you could be educating the world, right? through one lens. And I think that we w sh nothing should ever be forgotten that AI is programmed by humans. even if it is AI programming AI, there is a human there somewhere and that makes it behavioural.
Snehal: It is true. It is true. and that's what I would say like Instead of isolating AI in into any isolated use cases, it needs to be together with considering human factor and what is the best asset for any organization is having the trusted data, having a strong governance around it, having the shared business meaning, and and invest the same into cementing that along with the technology there. because the powerful your data is going to be the powerful AI can create the business outcomes for you. And that's the key. And and that this is going to help the leadership into the decision making, the right decision making for for their organization. This is going to be very powerful for their decision-making strategy and also to any initiatives that needs to start be started. This factor needs to be embedded at the beginning of any initiative.
Hugh Massie: Yeah, and it's with the right with the right investment. So I think AI can be very powerful and successful if all of these steps are followed and there's and but if one just treats AI as magic and that AI is going to get it right, that's also an even bigger problem. that that mindset's bigger bigger problem than the semantic debt problem, I think.
Snehal: Yeah, yeah, yeah. And And one of the examples I'll give you, I was doing my postgraduation program with the UT Austin, and we did the seven real world projects. In that the first most important factor is like curating your data, completely creating a very solid understanding of what your data is, and and then understanding your persona and then creating your AI product out of that. So as all throughout this our discussion as we spoke, it is very important that you have a trusted data, you have strong governance, you have shared business meaning, and your semantic debt is taken care of. It's gonna give you the best outcomes, best decision making for leadership. and it's and we are gonna thrive in this AI era where AI is so powerful that it can read everything and anything that you have in your organization. So it is very important that you have the powerful data to create a powerful AI products.
Hugh Massie: And I I think that's a great message, Snihal, to finish up on and I just want to thank you. You know, you are in my mind a world authority on data and organizational transformation and empowering AI in a way, you know, I think so far you're one of maybe the unsung heroes because everyone talks about, you know, anthropic, but what's underneath anthropic, you know, and you're one of the powerhouses that are going to drive the world forward the right way. I think that it can clearly only come from data. So thank you for participating, you know, with me today in the conversation. I really appreciate it and I think that our audience is going to get a lot out of this in terms of what they need to do and be thinking about.
Snehal: Thank you. I mean it was my pleasure and truly have a great pleasure talking to you. And this is these are really current topics. These are very important topics that people are trying to understand. And I've been following your and very much impressed with what you are doing at the DNA behavior. So it was truly a pleasure to talk about the behavioral science plus like understanding like the human factor and the data with it. So it was really great discussion with you. I I truly enjoyed it.
Hugh Massie: So Sneehal, just just to wrap up, where should people find you if they want to find out more about you and your work, what's the best place for them to go?
Snehal: so I have my website goldfriendconsult dot com. And my email address which I can provide you it is SNPall my first name sne-h a l dot mururkar the last name M-U-R-U-D-K-A-R at goldfriendconsult.com. so they can reach out to me there and I I'll definitely answer or help them in their data and AI needs for their enterprise level initiatives.
Hugh Massie: Fantastic. There'll be plenty of them.
Snehal: Yeah, great.
