[00:00] Debbie Reynolds: The personal views expressed by our podcast guests are their own and are not legal advice or official statements by their organizations.
[00:37] Hello, my name is Debbie Reynolds. They call me the Data Diva. This is the Data Diva Tosh privacy podcast where we discuss data privacy issues with industry leaders around the world with information that businesses need to know.
[00:49] Now, I have a very special guest on the show, a fellow Chicagoan, Rupal Dadu Patel.
[00:56] She is a Director of Enterprise Data and Analytics and I'm going to call her a data maven because I had so much fun talking with you about data. But tell me a bit about you.
[01:08] How does someone you know get into this data space and all the things that you're. You're working on? And actually that title doesn't really encompass everything that you do,
[01:17] so I love to hear that as well.
[01:20] Sure.
[01:20] Rupal Dadu Patel: I came into data organically, to be quite honest. Right.
[01:24] Every single position I held from when I started my professional career revolved around data.
[01:31] First, I was working for a big organization around supply chain and inventory management.
[01:38] We used edi. I don't know if anybody is familiar with edi, but that was back in the day.
[01:44] And then I worked with another big organization around data collection.
[01:48] Again, how are we collecting information about these various different companies?
[01:54] Those were the beginning of my data story.
[01:57] The biggest. As I evolved my professional career,
[02:03] it became very evident that the common theme across all of the data projects that I was working on revolved around the quality of the data and making sure that we really had a good understanding of what the data was,
[02:17] how it came into the organization,
[02:19] how it should be used,
[02:21] and how it's going to support the for whatever they needed to achieve. So that's the lens that I have focused on for the last, I want to say, 20 years around data management and making sure that foundation of data was going to help the organization use it to help deliver on what their objectives are.
[02:43] Debbie Reynolds: That's great. I'm glad that you gave that overview. Very cool.
[02:47] I want your thoughts on this, and I feel like you're probably the perfect person to ask this question. So I feel like every time there's like a new technology out,
[02:57] people they talk about like,
[02:59] oh, this is the holy grail, this is the new thing. And everything that they're doing it still relates to data. So no matter what application that you're doing, you need data, right?
[03:09] So some people, I've heard people, well,
[03:12] we don't need data strategy because we have AI, but it's like all technology uses data, period.
[03:19] So you can't be like, oh, the technology is going to take over and do all this stuff. But tell me your thoughts about this. This I find very humorous when people say this.
[03:28] But what are your thoughts?
[03:29] Rupal Dadu Patel: I cannot agree with you more that foundation is data. You can't do anything without data and people. Actually there's two sides of the equation, right? If you don't have the people side of the equation in play,
[03:44] you're not going to be delivering a good solution. If you do not have a good handle on what data you need for that solution and what it means to be a quality of nature security which touches on the privacy side of it,
[03:59] you're not going to have a good solution or you will have a solution that will have so many risk flags that you're going to have to risk mitigate and that's going to be end up being your job.
[04:12] So again,
[04:13] foundationally, we have to make sure the people side is really well understood, the process side is really well understood and the technology comes after the fact.
[04:26] Debbie Reynolds: I couldn't agree with you more on that for sure.
[04:29] I'm going to talk a little bit about data driven culture and building trust regardless of regulation. And this is something I talk about a lot.
[04:38] I at some point a couple of years ago, I literally stopped talking about regulation because it's like I do talks called beyond regulation. It's like there are certain things that you're just supposed to be doing anyway.
[04:51] Well, Dana. And if you just do those things, you'll be in a good place and you may have to do a couple little tweaks here and there, but I just feel like if you're managing your data the way that it should be managed, you're going to probably meet or exceed some of these requirements for regulation.
[05:10] But what are your thoughts?
[05:12] Rupal Dadu Patel: I couldn't agree. Again, your point taken. Right. The thing about data literacy,
[05:18] it really comes down to the business side of it. What is a business literacy? How about that? What as individual who works for that organization, do they understand what the business needs and why they need that data?
[05:30] If they understand that very well and they have that data ethics in mind. Because should all be thinking about data ethics,
[05:37] you are going to be designing and using the data in a very trusted way with privacy in mind and with the end goal in mind, which is really to support your organization and what they're trying to achieve.
[05:51] So for me, the biggest thing that anyone needs to be thinking about when they think about data is the business context in which that data needs to be used. If they don't understand their business and how their organization functions,
[06:06] then they can't. They're not equipped to use the data in a meaningful way.
[06:11] Debbie Reynolds: I agree with that.
[06:13] Let's talk just a little bit. I'm sure we're going to dabble a bit in AI, but this also makes me chuckle in a way as well.
[06:22] And this is companies you remember in the big data days, we're like, we just need more data, more data, more data. And it's like you have trash like you don't have anything that's going to help you get an insight.
[06:34] So people thought, oh, if I just hoard all this data and then somehow I'm going to get some meaningful insight out of that. But you really need to know what your end goal is so that you know what you need to be collecting in the first place.
[06:46] So I feel like we're having this conversation again with AI where people think we just need more data, just throw more stuff in and it's going to do magical things.
[06:54] But what are your thoughts?
[06:56] Rupal Dadu Patel: I use the word data rot a lot. There's a lot of data that organizations have and a majority of it say I would follow the 80, 20% rule where there's a lot of data that is redundant,
[07:09] right? It's duplicated across various different data stores and data processes, is obsolete. It is absolutely of no value to the organization at all. Even inferences and insight purposes. And it's trivial.
[07:23] The information that has been collected, you know, for whatever reason back in the day is no longer relevant for the business operations as it is. So I refer that to as data rot.
[07:34] And it's important to make sure that organizations are thinking about that data lifecycle management process.
[07:41] Especially because now we're moving into the cloud, right? We're moving into AI. We're moving into an area where storage of data isn't a problem.
[07:51] The processing power of that data is a problem. You are going to take your processing power dollars and you're just going to blow your budget if you don't really have a clear understanding of your data and the data life cycle management process.
[08:06] And Debbie, we are talking about privacy, right? So there are some state regulations that are coming in around retention and around making sure that organizations aren't holding onto data that is not Necessary.
[08:22] I believe California is going to be implementing a data delete process where you as a California resident can sign into a service and request that these organizations delete their data and they're going to be monitoring it.
[08:38] And a number of organizations didn't grow up in this modern data management era. Right. They've been around the block for a while. Hundred years of old organizations that didn't build their data foundations with data lifecycle management in mind.
[08:54] And they need to be revisiting that because regulations are changing and they're going to be implications if you don't adhere to that.
[09:02] Debbie Reynolds: That's true,
[09:03] very true. So I think also when you're talking about these older legacy companies that are pivoting or moving into or trying to modernize some of their operations, which I think totally they should be doing that.
[09:16] But I think what the privacy laws and regulations have done that, that we have not seen in the past is that it's forcing companies to get rid of things where there was no impetus for them to get rid of anything.
[09:32] Right. So like for example, like a data retention policy. A data retention policy tells you how long you're going to keep stuff.
[09:39] It doesn't force you to delete things. Right. So we're saying certain things. You really don't have a reason to keep this data for a longer term unless it ties to a purpose.
[09:51] But to me, that also connects to what you're saying about relating your data to a life cycle and related to a business case. Like why is it in terms of the business, why is it needed?
[10:04] Rupal Dadu Patel: Exactly. Yeah. And as you think about where the world is going, it's around knowledge management. Right. So it's not necessarily about where the data is stored. It's also about that business context layer that we need to be thinking about with the data.
[10:19] Where is it stored? What does it mean to the business unit? How can it be used?
[10:24] How does it need to be organized? Classified right for privacy purposes as well as for use case purposes. And what are we going to do with it and how are we going to use it?
[10:35] What is the use cases in which this data can be used?
[10:38] So what organizations needs to be really thinking about is data as a product and making sure that privacy is part of that component.
[10:49] And that is the way that our organization, not any organization, needs to move to.
[10:55] I strongly believe in that because if you don't have a clear understanding of your data products, you're not going to be able to use it for analytic purposes. You're not going to be able to use it for AI purposes, for customer retention purposes, for customer cross selling purposes, marketing.
[11:13] There's all sorts of different use cases that I'm thinking of. But data needs to be thought about as data products. And privacy has to be built into the foundation of how data products are developed.
[11:26] Debbie Reynolds: Yeah, I love that you said that. And I want to talk a little bit more about this. So I think once privacy became such a big issue, people thought about it as a reactive thing, like, okay, let's do all of our data,
[11:40] let's create all our data products and do all these cool things and then at the end think about privacy. But it really just doesn't work that way. It has to be thought of at the beginning and sort of baked into a process.
[11:53] So I like the way that you sort of think about it as a privacy lens as opposed to an attachment.
[12:01] What do you think?
[12:02] Rupal Dadu Patel: I couldn't agree with you more. I mean, it's really,
[12:05] this is around what data someone can see,
[12:08] what data someone can download, what data needs to be masked. Right. What data can be used for insight purposes,
[12:17] all of those things. And for the organizations that I've worked with,
[12:22] data privacy and access management has been a key component of how data needs to be looked at. Because if you get the data in the wrong hands, even though they signed an NDA, you know it is, it's still a risk to the organization.
[12:36] So really, really important to make sure we get that right.
[12:41] So that access management process is a huge component of the data. And when I talk about access management, I want to kind of talk through that a little bit. We focused our energy on access management from an employee standpoint and from a contractor standpoint.
[12:56] But now we need to be thinking about it from an AI model standpoint.
[13:01] Right. And if you don't.
[13:03] And AI is embedded now into so many applications and use cases.
[13:08] So we need to take the concepts of data governance and asset data assets and data products and then take that and add another layer onto it. And that is how are these data products being used by AI models?
[13:23] Right.
[13:24] And do we have the right roadblocks implemented to prevent them from doing certain things? Because now AI is also needs to be considered as part of our access models.
[13:37] Debbie Reynolds: Right.
[13:38] Rupal Dadu Patel: Especially when we're talking about Engentech workflows, things along those lines. And that is a gap that I don't think organizations are thinking of right now.
[13:47] Debbie Reynolds: Yeah. Let's talk a little bit about agents, AI agents and the gentic workflows. This frightens me to a great degree because a lot of what I'm seeing people do, first of all, they're really excited about it.
[14:01] So they're excited that they have these tools that can do these different things.
[14:05] But what I'm seeing is people advocating their own access that they've earned over the years to agents that don't know what they're doing.
[14:13] A lot of times when you're asking an agent to do something, it's pulling on other resources, right. So you don't know the changes that it's making in other places that may impact other departments.
[14:25] And then we have, of course in the news people saying, oh, we need to slow down because AI, we need to think about safety. But it's like it's a completely different world.
[14:35] And I think if we try to use some of the old ways to manage it, it's not going to work. Like we talked about.
[14:42] You talked about managing access based on employee or a contractor. But I've heard people say, well, we're going to make an agent like an employee.
[14:51] I said, but they are going to have more access on day one than any day one employee would ever have.
[14:59] And they're going to be doing things that you wouldn't allow even someone who's been with a company for many years to do. So to me, that's a totally different conversation.
[15:08] But what do you think?
[15:09] Rupal Dadu Patel: Yes, we do really need to be thinking through that.
[15:12] We have to design our access management processes to identify if it's an organ individual and then or a process or a system.
[15:24] And there needs to be different rules in place for that. And so you can't inherit is what I'm hearing you say, Debbie.
[15:32] Debbie Reynolds: Right.
[15:32] Rupal Dadu Patel: They're inheriting access based on who owns that data model or who's owning that ngintic workflow that they created. So like if you created the workflow, you're allowing your privileges to be inherited by that workflow.
[15:47] And this is where IT infrastructure and security needs to to come and identify these workflows and categorize them and organize them and not allow for that from a security standpoint.
[16:01] I don't know how to do it, but I think it definitely needs to be part of their playbook,
[16:06] you know, to make sure that we're thinking about it.
[16:09] Debbie Reynolds: Yeah, totally. I always talk about rules versus goals and I hear this over and over again in terms of how people misunderstand how AI is working in enterprises.
[16:20] So these agentic workflows things, they're goal oriented. So people, they get upset. Even like recently, this whole hugging face thing where I think OpenAI's agents went out and did some stuff with some other company and they.
[16:37] And these were not things that the company had sanctioned the AI to do. But if you give AI a goal,
[16:46] it's going to use whatever,
[16:47] whatever knowledge, whatever insight, whatever to do something. And they're not going to, you know, be thinking about laws and regulations or things that you told it not to do. All they're thinking about is the goal.
[16:58] But what do you think?
[17:00] Rupal Dadu Patel: Yeah, and it's not even about the laws and regulations. Right. It's about organization risk.
[17:06] Right. They're losing information that like int. They could potentially lose intellectual property that's very important to their goals and objectives. Right. It has nothing to do with PII or regulated data.
[17:19] It goes beyond this, in my opinion.
[17:22] So yeah, we have to be very cautious of what is going on with these workflows and gentec workflows.
[17:30] My stance on that. The biggest thing that we can do, what's in our control and what's in our power is education.
[17:37] We have to educate employees around the proper. Whenever you bring in a new solution, a new technology,
[17:46] there is this onboarding process.
[17:48] Right. And as part of that there you have to educate these individuals on how to use these tools.
[17:57] Create like almost like a checklist manifesto. Right. To say this is what you can do and this is what you can't do. Right. And have them use that as kind of like a guidance to design whatever they want to design.
[18:12] But education is one of the components that needs to be a part of it. And then again from a foundation standpoint,
[18:19] setting up these workflows or processes needs to be governed and secured by our IT security and infrastructure organization,
[18:28] part of the remit for them.
[18:30] Debbie Reynolds: I agree with that. I want your thoughts though. This is how I imagine things will change. So I think in the past when you onboarded an application, say let's say you had application for three or more years, like a minimum that you have,
[18:45] and the application maybe got updated every year, but not major, majorly.
[18:50] But now we're seeing a lot of these tools, especially things are in the cloud, especially things that are AI generated,
[18:57] changing rapidly, like very, very rapid changes in these applications. So it's almost like education has to be continual as opposed to just when you, you install application, because the application you bought six months ago is different six months later.
[19:16] So what do you think?
[19:18] Rupal Dadu Patel: Yeah, Exactly. So these SaaS applications used to have these quarterly releases, right? Or these releases that were bug releases and so forth. And as part of those releases there was a features list that was socialized to all their customers.
[19:35] And in my experience, I was a big salesforce person and we needed to decide and determine which feature we were going to accept and not accept.
[19:45] And as part of,
[19:46] I was a program manager for that in that situation. And we would send out like a business friendly impact,
[19:55] you know, story on what this means, what this new release means to their organization. So I think that's something that we need to do.
[20:03] But I think there's also that opportunity to make sure that vendors that we are bringing into our organization is aware that,
[20:13] that we control these releases in a way that we can manage because they shouldn't be able to update our ecosystem without our consent. Right. So it's almost like if we buy the solution as part of our contracts that we have with them, they are not able to update our instance of that application without the application owner's consent from within the organization because they're the front line for risk management purposes.
[20:44] Debbie Reynolds: Right.
[20:45] Rupal Dadu Patel: And I'm hoping, I'm not too familiar with how releases are being deployed right now, but I would hope that was part of the world.
[20:55] Debbie Reynolds: Well, I hope so too. I think a couple years ago there was a big uproar because one of the AI tools had done this update, some update they had pushed out to everybody.
[21:04] And for the enterprise customers they were like, well, the thing that you did like broke a lot of our workflow.
[21:12] You know, like you're building workflows on stuff and then they have a feature they stop using or change in some way that could create like a huge amount of disruption.
[21:20] So I think they had to roll that back and learn that, you know, even though it may be easy for them to push those updates out, as enterprises are really building workflows on top,
[21:30] they have to be very careful about.
[21:32] Because back in the day we used to have rollbacks and different things like this, like not just pushing things out without people knowing about it and making sure people understand what the impact could be to their.
[21:43] A workflow.
[21:44] Rupal Dadu Patel: Yes, that's right.
[21:46] So many things to think about and consider and, and it really just goes back to what are we trying to do and why are we trying to do it and then making sure you're putting the right roadblocks in place to, to kind of support that goal.
[22:00] Debbie Reynolds: Right? Yeah, yeah. What types of things are you seeing in your work sphere that's concerning you most about what organizations are doing right now?
[22:12] Rupal Dadu Patel: I think the biggest thing that I'm worried about is I don't think that we're in that we're educating our workforce around the AI capabilities and what they should be thinking about when they're using it.
[22:24] And this isn't only just about the organization, even about their own personal use of AI.
[22:30] Debbie Reynolds: Right.
[22:30] Rupal Dadu Patel: I don't think people are aware of what's out there. I think the other thing that I'm kind of concerned about around AI is losing our ability to think because we have these tools that will automate things for us and will help us write and articulate our thoughts in a clean and concise way.
[22:52] I think we're going to be losing our ability to connect the dots because we're using AI too heavily. And I'm worried about that.
[23:02] And,
[23:03] yeah, that's my biggest concern at this stage of the game.
[23:06] Debbie Reynolds: Yeah, I agree with that. I agree with that. Because critical thinking is really needed, especially when you're dealing with all these advanced technologies.
[23:15] I had someone on LinkedIn once say, well, all we need is,
[23:20] especially when we're implementing AI, all we need is a process where we can stop the AI at certain points and, like, interrogate it and ask for what it's doing.
[23:30] I said, well, the problem is people are having AI do things that they don't know how to do.
[23:37] So how do they know if they tell it to stop? They don't know how to tell it how to know that it was doing something wrong until it gets all the way to the end, you know what I'm saying?
[23:47] I think it'll be different. If the AI was doing something that you knew how to do,
[23:52] then you could tell it, okay, this is wrong. This is, you're going down the wrong path.
[23:56] We see people building stuff all the time that they don't know how to do. So how can you tell it to stop something that you don't know how to do?
[24:03] Rupal Dadu Patel: Exactly. And so that. And if you think about our education system, and I'm going way off here, but if you think about our education system and how we learn how to read and how we learn how to do math, we take things into character by character and character, put it together so we can kind of form how to pronounce a word,
[24:24] form how to pronounce a sentence, form how to pronounce a paragraph, we don't need to do that anymore. And that's going to really impact how we operate as the human in the loop,
[24:36] to your point.
[24:36] Debbie Reynolds: Right.
[24:37] Rupal Dadu Patel: And the same thing with math. If you had to do long division, you needed to know exactly each step that was taken to do that.
[24:45] And I think that is the same mentality that we need to use to build our AI workflows. Okay, put on paper what you're wanting to do.
[24:55] Tell me all the steps that we need to consider.
[24:58] You can use AI to help you with this. I'm not saying that you can't. It will help. But then at the very end,
[25:04] summarize it. Tell me exactly what you did. Document your thoughts. Document for me what you did.
[25:10] Because then as the human in the loop, you can say, oh, no, this isn't right, or no, this is not right. You're almost like becoming an internal auditor of what you're asking this AI agent to do for you.
[25:24] Debbie Reynolds: Right.
[25:25] Rupal Dadu Patel: But I think what's happening is there's this path of least resistance right now that's going on, and people are not doing that internal audit check.
[25:34] And that is so important. Before a process needs to be deployed,
[25:40] you have to do a QA assessment in it, right?
[25:43] Yeah. And similar going back to data privacy,
[25:46] organizations need to do a data privacy impact assessment. Right. To see how data is used and why it's used and what it's for.
[25:54] But again, it's for risk mitigation purposes. AI building AI processes is no different. To me, it's the same concept, different terminology.
[26:04] Debbie Reynolds: Yeah, I think that's true. I think maybe the issue is, first of all, the AI can work so rapidly,
[26:10] and I think people are very impressed, you know, as am I, with certain things that they can do and it does it really fast.
[26:17] But I think a lot of people don't take that initial step, like that interrogation step, that assessment step, that audit step, as you say,
[26:26] about what it's doing before you deploy it. A lot of it is like, well, let's just run, see what it does, and then,
[26:32] then look at it afterwards. And that's not a good thing because you may harm someone or harm something or.
[26:40] I think your point about intellectual property is a very good one, because I know we talk about privacy and we talk about people and human rights,
[26:49] but companies have assets that they don't want out on the Internet or they don't want it shown and may, like, give someone maybe a competitor an advantage or information that they didn't need to see.
[27:01] And we're seeing this also in like, mergers and acquisitions as well. So I think thinking about all those different ways that you use data and all of the harmful ways that you may not want data use, I think all that needs to, to come together.
[27:16] You've probably,
[27:17] as someone who's been in a data space for such a long time, how do you wrangle all these people and get them to do those preemptive or proactive things before they jump in to the deep end of the water.
[27:34] Rupal Dadu Patel: I think it's important for fun for a big initiative like this. Assessments are really important.
[27:40] And having a clear understanding of your inventory of data and models and processes is lost.
[27:49] Debbie Reynolds: Right?
[27:51] Rupal Dadu Patel: And so I think assessments is really important. I think making sure from an organization hierarchy standpoint, you have stakeholder buy in on why this is important,
[28:01] executive sponsorship, to say, hey, you know, we as an organization needs to move into the AI realm, right? We need to do this. But in order for us to do this, let's think about the foundation, let's think about what we have.
[28:14] And as part of that, there is this component of AI readiness that exists, right. And organ and not so the big organizations are probably already are doing it, right? I'm talking about organizations that don't have a clear picture of their process or systems or their data processes.
[28:32] So the first order of business is to be able to do an AI readiness assessment. And that AI readiness assessment is really thinking about what is a use case,
[28:42] what is a type of inputs that this use case needs,
[28:46] right? The people inputs, the data inputs, the system inputs, things along those lines and then not documenting them, but assessing them from a risk standpoint and then addressing for the gaps,
[29:00] right? And then keeping that engine moving. It's now being become operational.
[29:05] So you have this process to do this. And what this is going to do is not necessarily kind of pause the work. It's going to help you, inform you on how you can go about this the right way.
[29:17] You're not in remediation mode,
[29:20] right? Putting out the fire mode, right. That's the biggest thing in my opinion. We need executive sponsorship and you need to have a clear understanding of your processes across your assets and your AI components.
[29:34] And then again, all of it underlines is what type of data are you using and is that data relevant for this particular use case and for this AI? So you're basically codifying how data can be used for what you're trying to do, right?
[29:50] And if that data cannot be used for AI, then it's already noted and you know, you can put some roadblocks into place to prevent that data from being used for AI modeling purposes.
[30:01] Debbie Reynolds: I agree with that. I want your thoughts on how is privacy interjected into your data world when you're working. How does it get. Get entangled there somehow?
[30:13] Rupal Dadu Patel: So privacy gets entangled in. There's a lot. So for me, the privacy programs that I've Implemented included individual data rights, for example,
[30:21] and individual data rights, again falls under the concept of data products. Right.
[30:28] There are different Personas of an individual for an organization and you are going to want to manage their individual data rights based on that Persona.
[30:37] Debbie Reynolds: Right.
[30:38] Rupal Dadu Patel: So that gets embedded into the privacy individual. You're classifying your data based on the Persona as well as if you're a customer, you have PCI data. If they're a physician, not a physician, but like if they're a patient, you have HIPAA data, you know, things along a phi data.
[30:55] So that kind though that classification is a huge compone of identifying the data that you have.
[31:03] And then that data is also masked based on who's using it. So the access management layer is very important to me as an who has access to this data, what is their role within their organization,
[31:19] who can inherit privileges based on where they sit within the organization chart and who cannot. So privacy by design includes for me masking as well as classifications in order for that data product to be socialized for downstream needs.
[31:37] Debbie Reynolds: Right.
[31:38] Rupal Dadu Patel: So that's how we incorporate privacy in. And then we do data impact assessments. Right.
[31:44] As new data is being used for business functions,
[31:48] we assess where that data is going,
[31:51] who it's being shared with, things along those lines and then making sure that we have the proper controls in place for that.
[31:58] Debbie Reynolds: I feel like companies, I want your thoughts. I think they understand data, how to manage it when it comes in,
[32:07] but somehow they lose track of it as it goes through the data life cycle. And this is where I see a lot of privacy issues, where for example, like let's say patient data was supposed to be used for something for patient purposes, but then the marketing department gets a hold of it and then they say,
[32:26] oh, let's market this to this person. You're thinking, no, no, no, you can't do that. The reason why you have the data and the purpose somehow can get lost.
[32:35] And this happens a lot in organizations. But what are your thoughts about that? Because I know you talked about data life cycle, which I love talking about that because a lot of times people think about the beginning of data capture, not like the life of it all the way through.
[32:50] Rupal Dadu Patel: So this is again another reason why that concept of that knowledge management is so important.
[32:56] And one of the organizations that I worked in, we did was we created a 360 view of our data assets. And what that meant was what is a dictionary, what is the business terms based on this?
[33:09] Where's the data stored,
[33:11] where is it coming from and how is it evolving as it moves from point A to point B, the data lineage component of it,
[33:18] the data classification of that data.
[33:21] Debbie Reynolds: Right.
[33:22] Rupal Dadu Patel: And then this is the part that I think you're trying to get at is what business unit, what business capabilities should have access to this data.
[33:32] Debbie Reynolds: Right.
[33:33] Rupal Dadu Patel: That is what you're talking about. And if they shouldn't have access to it,
[33:37] then it's already written there. And again, if you bring that think the term,
[33:41] I know the term is that technology and the ontology of this and you put that layer in to your overarching data management foundation,
[33:51] then you will prevent these data uses that are not approved.
[33:57] So again, it's that building that foundation, and we have to build that foundation with that privacy in mind.
[34:04] Right. And so the org chart is a really big component of how to do this.
[34:11] Right. An organization has their org charts. They know,
[34:14] you know, there's a profit and loss centers and things along those lines. Every employee is linked to one. You know, you bring that into the equation as you're defining this from an operation standpoint.
[34:26] Yeah,
[34:27] I've done it.
[34:28] Debbie Reynolds: Well, that's a very wise way to do it. So I'm glad you were able to lay that out. It's very important. I feel like sometimes organizations may think that you're slowing a process down, but actually you're helping them, enabling them to have speed once they pick up the pace.
[34:45] Because then you're not reinventing the wheel every time you pull in a new, you know, data set or you every time you pull in a new application because you exercise that muscle throughout your initial work and then you're just applying the things you've already learned to this data.
[35:03] Rupal Dadu Patel: Exactly. Yeah. So for organizations that have mergers and acquisitions,
[35:07] you put that framework together, you put those processes and engines in place to bring that data into the organization is just tenfold faster.
[35:17] Debbie Reynolds: Right.
[35:17] Rupal Dadu Patel: Whereas you're reinventing the wheel every single time. And that's not necessary.
[35:23] Same thing with onboarding an individual employee. Organization structures change.
[35:28] Right. So you can bring on an employee and get them up and running so much faster if you have this data foundation in place.
[35:37] Debbie Reynolds: So true. I agree with that.
[35:39] So, Rupal, if we were the world according to you and we did everything you said, what would be your wish for privacy anywhere in the world, whether that be human behavior,
[35:50] technology, or regulation?
[35:54] Rupal Dadu Patel: Gosh, I think I love the concept of regulations. Regulations are ever changing and evolving.
[36:00] We need to build data privacy into the ethos of the organization that data ethics is the same thing as being a human.
[36:11] Debbie Reynolds: Right.
[36:12] Rupal Dadu Patel: And so if I was that person,
[36:14] what is it you want to do onto others as you want it to do onto you?
[36:19] It's the same thing with data, right? When you could do it, when you should use it, you know, when you shouldn't use it, you know, and just having that ingrained into individuals is just so important because if your data gets into the wrong hands, how would you feel?
[36:36] Debbie Reynolds: Right.
[36:37] Rupal Dadu Patel: I think data ethics needs to be part of the ethos of an organization, and I would work really hard to build that wherever I am now.
[36:46] Debbie Reynolds: That's incredible. Well, thank you so much for being on the show. I absolutely love this episode. I love chatting with you.
[36:53] You have such a amazing brain around data, and I feel like that's you're filling a gap. There's a void that's definitely needed within organizations that's really taking a more, like you said, a 360 or more holistic view of data within an organization.
[37:08] And you're really solving problems before they occur, which I think is amazing.
[37:13] Rupal Dadu Patel: Thank you. Thank you, Debbie, for bringing me into this conversation. I really enjoyed it. I loved your insight too, because I feel like you and I,
[37:22] you know, are talking the same way. I love it. We're like minded individuals when it comes to data.
[37:27] Debbie Reynolds: We are. They call me the data diva for a good reason, for sure. Oh, my gosh.
[37:32] Well, thank you so much and I really appreciate you being on the show. This is tremendous.
[37:37] Rupal Dadu Patel: Thanks again.
[37:38] Debbie Reynolds: All right, talk to you soon.
[37:40] Rupal Dadu Patel: Bye.
[37:41] Debbie Reynolds: Bye. Bye.
