¶ Intro / Opening
On this episode of data driven Frank and Andy interview Lauren Mafayo author of Designing Data Governance from the Ground Up Data governance has become more pressing of late, what with all the advancements in generative AI systems. Tune in for a fascinating look at data governances, civic technology, and more. You. Hello, and welcome to Data Driven Podcast. We cover the emergent fields of data science, AI, and machine learning. Today,
I'm here with Andy. My voice is a little crackly because of a sinus infection, but it's all good. I've gotten on the meds and I am definitely feeling like I'm on the mend. How are you doing, Andy? I'm well, Frank. And I just heard how you were doing. Actually, I knew a little bit about it because you texted me when you were in the throes of it, and I knew something was up because usually you communicate
more. I was like, Frank's down for the weekend. And I know you've been having very busy weekends the past little bit for something that people will know more about later, right? Much later, probably. But it's all good. It is all good so far. It's ended well. So for folks that we're going to release this episode, we're recording this on July 17, we're going to release this probably on July 18. And you'll hear me refer to a legal case. It looks like that will be resolved this
week, hopefully in one form or the other, and it's gone our way. That's all I can say right now. But it is good news. Speaking of good news, we have with us an excellent guest who's based in the DC area. So not that far from Chateau Lavinia. It is Lauren. Sorry, she will correct me, but she's a published author. Her book just came out talking about designing data governance, which is a topic that just more and more keeps coming up. And I think that if you're a data engineer and you think
I don't have to worry about that hold up. Maybe you should need to worry about that. Even data scientists? Especially data scientists, I would say, and doubly so if you're in the generative AI space. I think we'll see what we get into that. And she has a very interesting background, so I'll let her explain it. Welcome to the show, Lauren. Thank you guys, for having me. I'm really excited to be here and to chat with you all. Yeah, likewise, likewise. So your background is
amazing. You studied overseas at Cambridge, I think. At LSE and the London School of. Economics, which is like, wow, I half expected you to have a British accent, honestly, because I wasn't sure. And you also have spent some time doing arts and design, so I found that fascinating too. I actually am a service designer in my day job, and so I work very closely with data scientists and engineers to design things like pipelines, cloud architecture, environments, different service models for
chief Data Officers. And so I always say as a service designer that I'm the user advocate on a project. I'm the person who is tasked with helping the client define who their key user groups are. And once I do that, I conduct user interviews with people who fit those demographics to figure out what they like or dislike about a product or service. I capture the results of those interviews and design assets like personas and journey
maps. And then ultimately I do work with people like you, data architects, engineers, scientists to build a product that will hopefully solve the pain points that we uncovered in the user research. Fascinating. And you were in the Civic Tech space if memory serves as well, which is a fascinating space that once upon a time
I was on the Microsoft Civic Tech team. Yes, I am. So I work for an organization called Steampunk and we're a human centered design firm that builds solutions for federal government agencies because as we all know, the federal government is the most progressive when it comes to tech and so they barely need us at all. But the reality actually is that they need us quite a bit and that we very often come in and have that human centered approach that many of their tools
were just not built with. And so then we come in and often try to improve them and improve the user experience. And user experience in that context is really about getting the right services to the American public, which I
think is what makes the work so interesting. It's not commercial products, it's things like improving unemployment benefits and how easily it is for people to, how easy it is for people to access them, improving the ease with which you can send folks overseas in official roles, defining the service offerings that a Chief Data officer is going to provide its colleagues. And so the problems that you solve in Civic Tech I think
are really fascinating. And I think COVID was the final confirmation that all of these systems are long overdue for major upgrades which we are seeing the influx of now. Yeah, you don't have kind of good user design or good user experience as part of the RFP that went out for building these large federal systems. That made was probably not a bullet point on the list, not at worse. So for those not familiar with Civic Tech, how would you define it? I would define
¶ Civic Tech serves the public through technology.
Civic Tech as technology which exists to serve the public. And the public is very broad. I would define the public further by saying it's citizens of any country or area where the tech exists. And so for instance, Civic Tech encompasses the tech in a town that my hometown, for instance, NATIC, Massachusetts might use to serve residents of NATIC. So this could be anything from tech that allows people to pay their bills online to applying for benefits. And then likewise I
work as a designer in the federal space. And so I work with US federal agencies to improve the way that they deliver services to the American public. And the public in this case, is any American who needs to use those services. But then we get more granular about who those particular user groups are. So, for instance, I have worked on many projects in the past with the Department of Agriculture, and within the Department of Agriculture there are many different
subdivisions that serve different user groups. And so then I will work with my client to define what those user groups are and figure out how we can tailor a user experience and a product to meet those unique needs. But I would broadly define civic tech as any technology which serves the public. And the public can then be further defined into groups based on things like geography, but also things like role, the day to day experience,
things like that. That's a good definition because it used to be very nebulous in terms of what it meant and the implications thereof. But I like your definition. It's probably the most cogent I've heard to date of the field. Thank you. Now this explains so how did you get into data governance, right? Because this is something well, let's start before we do that. How would you define data governance? I love the fact that you
¶ Data governance: a holistic, cultural business strategy.
start the conversation by asking me to define it, because I think like many terms in tech, it is often left undefined. And that's why there's not only a lot of confusion about it, but also a lot of resistance to it. I think people have in their heads that governance is purely compliance and that it is a blocker to innovation and to tinkering. Other people think that it is something that you can quote unquote, ship after
deployment. And I have had C suite leaders say as much. They've said things like, we'll do data governance later, or we will deliver it in the next contract after production. And that refrain is still unfortunately common. So I define data governance as the strategy you have to encompass the people, processes and tools that help you manage your data at scale. And I often say manage your big data at scale. Big data, as we know, is another buzzword that often means both everything
and nothing. But I use big data in this context because the reality is that most organizations have more data that they both ingest and produce than ever before. It is too big for one person or one team to manage on their own. And that's why you do need this holistic data governance strategy that is really a business strategy before a technical strategy. Your data governance should never be divorced from what you're doing in development and production environments. It should be
integrated into those environments. But at the same time, I think people make a mistake when they think of data governance not just as pure compliance, but also purely as a technical problem to solve. Because the more complicated reality is that it's a cultural transformation that your organization needs to be invested in from the top down. And that's really how you gain success from data governance. Now, that's a good way to put it.
And that's why I wanted to define it, because it doesn't have a very firm definition, right. My definition, that my operating definition is pretty close to yours. I'll say it's really because in my day job at Red Hat is like they ask, well, what does your product do for data governance? And I kind of laugh and say, well, not really much, because data governance is largely around, yes, it's people, processes and technology. But 80% of that is nothing is not technology. Right.
And you need a vehicle to make it happen in the technology space. But the people in process part, those are going to be the hard ones. Absolutely. And that's why it is so tricky. I think it's also why relatively few organizations have made a lot of headway. And that's also why I think it's really important to frame data governance as a cultural transformation that you can design and embed into your business strategy. You really cannot
separate the two. I think a lot of people have been saying that for quite some time now, but we're really seeing the results of that and rather the results of not doing that now we are in a pseudo recession, if not an actual recession. Tech organizations have certainly been acting like there's a recession with both layoffs of employees, but also in their buying behaviors and in not buying as many cloud tools and
pieces of software that they used to. And so it's more important than ever that whatever technology you're investing in is producing tangible outputs for your organization. And so we're seeing the consequence of trying to divorce data governance from your business strategy. It's just no longer an option to separate the two. No, I totally agree. And Andy looks like he has a question, but I want to get this out
there. I think part of it is that a lot of organizations, and I mean legacy organizations probably, I would say federal, it would definitely fall on this, is that it's only been in the recent years, maybe decade, that we've thought of data as an asset as opposed to a byproduct of some other process. And maybe that's it now it's
something of value. And as with anything of value, you probably should have processes not guards around it, but gatekeepers or gates around it just to make sure it's not wasted, it's not contaminated, that sort of thing. That's where my head is at.
¶ Data as tangible asset, managing as product.
I agree with that. I think data as an actual tangible asset is a relatively new concept, certainly within the last decade. And I think what's also new about it is the pure volume of data that exists in the world today, there is more data produced and ingested than ever before, and that number is
certainly not going to go down. When you think about all of the Internet connected devices that exist, when you think about the explosion of remote work and the fact that now employees are doing work for their organizations on private devices, which means that you can be having organizational data that exists in several locations,
which is a very tangible reality. And then I think that lends itself to the broader conversation that I see happening in data circles now about managing data more as a product and less as a service, which is an approach that I largely support because a big part of what you need to do to be successful at data governance is defining clear data domains and subdomains within
your organization. These are the key areas that your organization collects data on, and then it gives you a way of categorizing them more clearly, rolling them up to specific owners. These would be equivalent to your product managers if we're using the product analogy. So there's a lot being done to reframe big data in this way as an
asset that you manage like a product. And I think there's a lot of value to that, rather than the top down data as a service model that begins and ends with it and begins and ends with people who really lack the context to make those decisions about data and its quality across domains, I. Think that's really
important. Lauren and what would you say to an enterprise or just maybe a small to medium sized company that says, yeah, we understand all of that and they kind of give mental assent to it, but they think about their culture and the way they've always done things and they can't bridge that gap? That's a great question because I
¶ Implementing data governance: start small, connect to business.
think that is realistically. Where the biggest blockers occur, people are messy, they're intangible, they all have different motivations, even if they work for the same organization, they not only have different roles, but they have different end goals. Very often you have people in organizations who do not want change, they want things to say the same, they have a vested interest in it, even if that is arguably not what is best for the organization in
the long run. You will have people who are invested in not changing the status quo, especially as it pertains to data. I think a lot of that comes down to the fact that data governance has not been practiced to the degree that it should have. And so when people look at how much data they have in an organization and then they think about not only the work it would take to create data governance standards from scratch, but then to retroactively apply those standards to the data they have, it gets
very overwhelming very quickly. And so what I would say to someone who is on the fence about implementing data governance is to start small. To start by looking at the key data domains in your organization. So these are the areas like sales Data, marketing data, customer success data, where your organization is producing and or ingesting data about from a high level. I would also tell them to start small by not only defining those key data domains and
respective subdomains. For instance, you could have a data domain on sales data and then two subdomains could be inbound and outbound leads and those are two subdomains you can collect data on. But then you also want to apply that data to a particular project that is contained and that has been already greenlit by the sea level leadership as having high value to the organization. I think
that does two things. It helps you contain your efforts so that you are not reinventing the wheel across all areas of the organization, and it also ensures that you are working on something that senior leadership really cares about that is also essential. I talk in the book about finding the right sponsor for your data governance efforts, and that really is crucial because like any big
transformation, it has to be a top down effort. If you're the Chief Data officer and your C suite, your chief executive officer is not on board with data governance, you can make some progress. Because, again, if you're a senior data leader, your entire job is to strategically manage data as an asset. And so you can make some progress. But without that high level buy in and without connecting your efforts back to the business, you're really going to stall. So I would say start
small. Look for a strategic project where data governance can add value, and then do everything you possibly can to connect your governance efforts back to that business goal. So it sounds like someone should write a book about doing data governance from scratch or something like that. That would be a nice idea. It would have helped me on some of my early projects, which is why I wrote the book that's well, I. Was
going to lead into that. And you mentioned the book in your answer, and Lauren has written a book for those who are listening, and it's called Designing Data Governance from the Ground Up. And I just picked up the ebook. We were looking at your bio before the show. Frank and I connect about five or six minutes before the show, and I said, that sounds like something I need to dig into. So I picked it up, I'll read it. I've got a little bit of vacation coming up here starting at
the end of the month, so maybe I'll get to it then. I'm looking forward. Hopefully you'll read it on the plane there or back. Because I always joke that if someone's reading my book on a beach somewhere, something's gone wrong, because this is not exactly a light hearted beach read. And I always joke with people because when I encounter resistance to the concept of data governance, I joke with them, well, you might not want to read my book, but you're going
to have to read the book at some point. So hopefully it will be helpful when you do. I look forward to it. And as we were talking a little in the virtual green room about this, and I said, I'm basically a data engineer. I came into data from software and I made the leap about probably 20 to 25 years ago when a lot of I would call it process control, because before I did software, I was in manufacturing. So it had a lot of the same types of thinking around engineering and process control.
And even back then, some of the buzzwords that sound new in software are new ish we were doing in the 90s in manufacturing stuff like Kanban and Six Sigma and those sorts of metrics collection. And I was very fortunate to be trained by someone who was trained by W. Edwards Deming himself on that information. So very fresh, probably some insights that I'll never share, but just interesting to
get. Definitely a true believer and someone who came at it with an open mind and really understood it, but these sorts of things that have grown out of that, and I see this as growing out of the data governance is one of the things that grew out of a combination of compliance and quality. Would you agree with that or would you correct me? No, I do agree with
¶ Data growth, lack of management, legislative progress. Clear framework for data quality needed.
that. I think that actually hits the nail on the head. We have let data grow unchecked, broadly speaking, and that is because we just didn't know, as an industry and society how to manage it. You're exactly right that there are people who have been data architects, engineers, scientists for decades, and they've been doing this work for a very long time outside of
the public view. But what's different about the work today is the volume of data that is produced by consumer products and the amount of sensitive data that is effectively floating out in the world today through various cloud systems and various products that are used. And to that end, we're now in the earliest stages of figuring out how to manage that from legislative standpoints, both in the US. And abroad. GDPR legislation in
Europe comes to mind. That's fairly recent legislation that gives EU citizens a lot more personal rights over their personal data and what organizations can do in terms of profiting from that data. We do not have the equivalent of federal legislation here in the US. But I do see that changing over the next five to ten years. And I think what you also said about
quality really rings true. That's a huge issue because we as an industry really lack consistent, clear standards which define what data quality is and how we should be measuring it. And that's a big difference. If you look at fields like medicine law areas that have very high impact on the public, they have pretty clear governing bodies and standards for how doctors and lawyers should do their
work. We have things like IEEE, we have the association for Computing Machinery, we certainly have membership organizations where people can get together and discuss these things and debate these issues. But we really lack a clear framework for data quality and compliance, which I think is very long overdue. So I do see that as being the double pronged issue today. And I'm also curious what your take is, as someone who's been doing this work for
decades. How have you seen data governance evolve from the 90s through to the present day? Well, it's interesting as I've made the transition from being an employee to being a consultant, which happened around 2005, 2006, I definitely saw some difference there. But as an employee at one place, and actually I was a contractor there too, attempt they worked with medical devices. And so there I saw a strict compliance, but it almost fed down
from the culture. You mentioned culture earlier as being very important. I totally agree. But it was almost an accidental culture shift that came from the medical device part, the medical part of the medical device field into all aspects of software and data. And it was really interesting to see how that sort of thinking led to almost a practice of data governance. And we weren't even calling it calling it data governance back then, right? We were just considering it software and data. That was
all. I fell under that umbrella. And having that experience there was very eye opening and going from there to more of a startup culture, which not picking on startups. There's a priority difference, though, between that and somebody in kind of a more stayed and stable environment. And I'm not picking again, I'm not calling startups unstable. There's a lot of benefits to startups and a lot of innovative cultures, and some of that wasn't
present in the more medical device environment. Some of the benefits of that kind of drive and ambition and go, go and get things done. But it's very easy to overlook. And I saw it, I saw important aspects of what we now call data governance and really just good engineering practices. Some of that was overlooked, some of it was deprioritized for what I consider to be mostly legitimate business concerns in a startup world. I would agree with that. I think when you
¶ Startups prioritize innovation for survival. Large industries restrict innovation due to regulation. Motivations and context are key in governance.
consider startups and the landscape they're in, they have to innovate and be different or else they will not survive in the marketplace. And so their priority really is to move fast and figure it out later. I gave a talk at Data Architecture Online last week and the keynote moderator made a joke about how developers are often like, don't bother me with requirements on coding, meaning they're tinkering and they'll figure it out
later. And we've really taken that approach with data and that it's a really tricky balance to balance those standards and the creation of those standards with the need to innovate and stay in business. And that's really what startups are focused
on. And then on the flip side, you have these large, highly regulated, highly bureaucratic industries like government, healthcare, medicine, law, which are highly regulated, and they have to exist to be stable and to provide services in a way that their users can rely
on. And so innovating, not only is it not the priority in those environments very often, it's also an inherent risk because people in those environments are not really rewarded for doing something in a new way, but they will be very highly penalized if something goes wrong. I think you talked and touched on motivation earlier, and you really have to examine the motivations of whomever
you're working with and consider the context. The book that I wrote is a 100 page six step guide to designing your first data governance program from scratch. And it is short enough because there is a lot of nuance when it comes to data governance. When you implement a data governance program for 100,000 person multinational firm, that is going to look very different than doing it for a 25 person startup. But the key aspects of governance are the same,
I argue, across those nuances. And so that's why the book is short in the first instance, because it's meant to be the first prelude to whatever gets more specific about how to do data governance in your own environment. And that context per environment is really crucial. No, I mean, that's a good point. Data governance, it's come up more and more in my day job as well, because it becomes and it's
also interesting. And as the world's imagination is captured by generative AI, I think it's important to realize the generative AI. Well, first off, there's a lot of legal questions that remain unresolved, right? Like, if I tell it to produce a novel in the style of a particular author, andy's laughing because we've been doing some experiments with that. I was muted, but I was laughing. You were laughing. Yeah, more on that later. But no, I mean,
what does that mean? If you produce an image in the style of a particular artist, obviously, that is but I think the legislative hammer is coming down on that. And my opinion is it's probably best to start with governance today to save you what a stitch in time will save nine legal bills later. Like something like that. Do you think that generative AI is really going to make the data governance cool, for lack of a better
term? That's a really interesting question. I think it is absolutely going to make
¶ Data governance and copyright infringement in generative AI. The future of consumer rights and cybersecurity.
data governance essential. And I was speaking to somebody on a separate podcast this month about this very issue because you mentioned writing a book in the style of a particular author giving generative AI the prompt to write a novella in the style of cormac McCarthy, for example. In that case, you are maybe not copying or plagiarizing cormac McCarthy's work directly,
or maybe you are. It really depends on whether the generative AI can actually understand what you mean, and it can understand cormac McCarthy's style of writing enough to produce a novella in his likeness, if you will. Likeness is a very interesting concept, I think, these days. And you're right, it is incredibly murky from the legal standpoint. And I was speaking on a podcast recently about this in the sense of where when we look at the legal landscape of generative AI, where
is there going to be progress? And rather than making progress on the consumer data privacy and consumer rights aspect of the issue, I actually think that we're going to see more progress made and more cases brought to court on the grounds of copyright infringement. If you look at things like using a music in a movie or using images that a corporation owns in a book, I just went through this with my own book. I wanted to use commercial software to make a few diagrams
and use templates to do it. And my editor said, are those templates that are pre built into the software? I said, yes. And he said, you either have to get permission legally from their legal department to use those in the book, or you have to create some from scratch and make them yourself. So I chose
the latter because it was the path of least resistance. And I think when we consider generative AI and what that means for data, we in the United States are going to see more progress on the grounds of copyright infringement than we are on data privacy and consumer rights in the short term. Now, having said that, I think humans are
inherently reactive. And I do foresee in the future, within the next five years, certainly there's going to be a data breach to such a degree that there is going to be enough groundswell for organizations to really get serious about protecting consumer rights and as it pertains to data. The other model you can look at is what's happened in cybersecurity three to five years ago. There were very few conversations happening about being proactive when it comes to
cybersecurity. And in recent years, we've seen a large increase in breaches, not just within software companies, not just within organizations, but even breaches of oil and gas pipelines, things like that. And so just like with data governance no longer being a nice to have, it never was to begin with, but now it really is something that you need. Likewise, we're seeing tech teams really prioritize cyber, not just in their pipelines, not just on the technical side, but
also creating a more cyber literate workforce. And. I think there's actually a lot that data practitioners can learn from their counterparts in Sizzos to drive the needle on that front. No, that's a good point. I think connecting those dots are important because when the C suite realizes that this isn't a game anymore, when the SCADA drivers got hacked, or when the Colonial pipeline incident happened, I think that realized in obviously a number of ransomware
attacks. I think security became very serious, like, oh, wait a minute, this could affect us and it's not optional anymore, or nice to have. Right. And I think data governance is going to follow that same thing. I think that's an interesting take that you have, is that up till now, the only driver in this space has effectively been privacy legislation, right. GDPR probably being the poster child for
that. But I can easily see fear of being involved in some massive copyright lawsuit would probably like, I know there's some controversy about how GPT was trained, right? Like he was trained on Twitter data and then Elon Musk said, wait a minute, did you get anyone's approval for that? On that
¶ Encourage caution with sharing proprietary information
note, I would also encourage people because every now and then I have the strong urge when I am transcribing, for instance, user interviews, to use a tool like chat GPT. It would be incredible if I could feed that video content into a system to spit out an accurate transcript. And that is absolutely not an option for the role that I'm in, for the industry I'm in. I cannot give that proprietary
information to anyone outside of my organization. And if I did, the consequences would be things that I don't even really want to think about because I am beholden to keeping that information private. And what that calls to mind is the Samsung incident. Pretty early on in Chat GPT where folks fed proprietary Samsung data to chat GPT.
OpenAI owns that now. Again, we as a society, we as an industry don't have the full context or real comprehension of what that actually means, what ownership really means. But on a very practical level, it does mean that highly sensitive commercial data is now with the hands of this very large nonprofit to be used in very different contexts in very different ways. And the consequences of that are really going to be felt and continue to be felt, I think, over the next several years.
That's interesting. I was just going to say it's almost like the I'm not sure how accurate it is, but knowing the source I heard it from, it's probably likely that a game manufacturer received proprietary information from a defense contractor in the US. I don't want to get too specific. It sounds like something is hitting the fan and it's not parmesan cheese. Well, it
was an argument. The bit that I will share is it was an argument about someone had made a guess about what the interior of some piece of equipment looked like and someone said, no, it looks like this. And they actually supplied documents to prove that. And that wasn't good. Wow. Yeah, that was pretty wild. It was like all on discord server too. Exactly. Which was notoriously secure. So many wrong things about that, yet that happened. It's off the charts. But I mean, it's a good example of good
intentions going horribly wrong. And you think that's a thing in data governance as well, like a risk?
¶ Bias in AI and data governance intertwined. Risk reduction, troubleshooting. Not all intent is negative. Challenges in data work solvable. Nonprofits and cybersecurity models for governance.
Absolutely. And when I talk about bias in AI, which is one, I don't believe, again, that data governance is separate from bias mitigation in the training process. I think data governance is a form of risk reduction and bias troubleshooting. And I do think that the overarching issue here is that we really need to think of this as an integrated problem that is one with the business. But I also think that people it's a misnomer to say, of course hackers have nefarious intent in many
cases. Of course, there are always going to be people that want to manipulate data, that want to use it to cause harm. There's no doubt about that. But the vast majority of times when we see the biased outputs of algorithms or we see data governance gone wrong, no one was trying to harm someone. There was no negative intent. There are many complicated technical reasons why an algorithm can produce biased outputs towards one user group over
another. And this is kind of where when people say, assume positive intent, I think that only goes so far because I don't believe that most developers or data scientists are trying to or executives are trying to harm people by
a long shot. They're really doing the best that they can. But if the end result is still that people's rights are being abused, that resumes are getting screened out automatically instead of being given the proper consideration, if those negative results are still occurring, the intent, how much does it matter? But I do think that's an important distinction. Rather than painting the industry overall as a group of bad people with ill intent, I just don't think that's accurate, and I think there's
a lot more nuance to it. It's also important, I think, to show that while these challenges are part of the job, they're inherent in the work of doing data today. Whether you're an engineer, a scientist, a governance person, this is part of the job. And so to that degree, it's somewhat inevitable, but it's not unsolvable. There are tactics that you can use to improve your work in this space, and so I don't want it
to be a doom and gloom scenario. There are things that we can do as practitioners to avoid a lot of the consequences we're talking about, and there are a lot of blueprints out there for how to do this. Like I mentioned, cybersecurity is doing a lot to educate workforces on how to spot phishing attacks.
Things like that if you look at it, governance from a stewardship perspective and a governance council perspective, if you've ever certified on a nonprofit board, nonprofits are actually surprisingly advanced when it comes to things like data governance. When I was writing the book, I found many universities washington University in St. Louis comes to mind that have full websites devoted to their data governance charter, who serves on the governance
council, what they manage on it. And I'm sure those people would tell you that their governance council is far from perfect, but they're doing the work, they're holding themselves accountable, and they've set up the structure to succeed. So nonprofits and the cyberspace are both two really strong models to look towards when we're thinking about what the future of data governance looks like.
No, that's a good way to look at it. It's an evolving field, and it's interesting how it's finally coming up, and it's becoming more and more prevalent, at least in the conversations I have. And
¶ Encouraging shift in conversation about data governance.
that's encouraging to hear, because like I said, when I was pitching the book and then writing it, I felt confident that this information was necessary, that people in the field could use it. But at the same time, I was seeing relatively little being written about data governance. I was seeing a lot of articles on different things you could do with data from the data science side or engineering side, but I wasn't seeing a lot about governance, and there was that nagging part of me that
worried. I feel confident about this book and its subject, and I do worry that it's going to land with a bit of a little thump and then go nowhere. But I've actually really seen the conversation in our industry shift this year. I think it's no accident that that happened when Chat GBT became mainstream, when Generative AI officially became mainstream. And that really was my thought all along, was that we were going to reach a
tipping point where data governance was necessary. And so I would even go so far as to say when the book was in beta last fall, I still had some of those concerns about whether it was going to be relevant enough or perceived to be relevant enough, and I don't have that doubt anymore. So it's interesting. I see that there's an Audible version too. That's
awesome. There is. And so they did turn it into an audiobook. So if people want to read it, they can either pick up an e copy, which is available on any ereader, they can also order a print copy, but it is also available on audiobooks. So if people want to utilize that I know that audiobooks are preferred for people on the go. I listen to them at the gym or on planes, and so I find that audiobooks can be a great
alternative. If you don't have that time to sit and read every day, you probably at least are sitting down at some point during the day, whether on a commute, whether on a plane. And so hopefully the audiobook can help. No, absolutely. Because of circumstances related to what I mentioned early in the show about the good news, I was just spending a lot of time in the car between here and Pittsburgh. So I've gotten a lot of audio books done in there and I think this is an
awesome conversation. This could probably go on for the 2 hours, but I want to switch to the pre canned questions. But while hopefully Lauren, you've had a chance to review those before. Oh, Andy just posted them, it looks like. Well, let me post them over here in our team's chat. Oh, I just did it. They're not brain teasers, but they're just fun little questions that we have, we ask of every guest.
But I will point out that Audible is a sponsor of Data Driven, and if you go to thedatadedrivenbook.com, you could pick up a free book. And I'm looking forward to listening to your book. Lauren. Awesome. Thank you so much. That really means excellent. Yes. And if listeners want to buy the book, you can go to Pragueprog.com. That's Pragprog.com. The book is called Designing Data Governance from the Ground Up, and your listeners can use the code Datagov 23 all Caps to get 35% off the e copy.
So if folks are interested and they need a little bit of a boost, that code should be good, and I would love to know what folks think. So I'm happy to be connected with on LinkedIn and if folks want to leave reviews of the book on sites like Amazon and Goodreads, that is also hugely helpful. Those reviews really do make a difference in books getting found and discovered on those platforms, so every review helps. Awesome. All right, our first question. How did you find
your way into Data? Did you find Data or did Data find you?
¶ Data found me, sparked interest in AI.
Data did find me. I'm a writer at heart, and I have a background in mixed methods research, journalism, and digital media and content management. I started using open source CMS systems to manage that content. So that's my first foray into open source tech and communities. But I didn't really get interested in Data until I was a research analyst at Gartner and I started learning about AI
that way. That's where I started hearing about different types of AI, things like natural language processing versus robotic process automation and how you could use these different types of tech to
solve very specific business problems. And I was surprised by how interesting I found that whole aspect of it and how interesting I found the fact that at the end of the day, AI is data, and the more you learn about data and the more you know about it, the more you can use those technologies effectively. Awesome. You want to take the next question, Andy? Yes, sure. Sorry. I was thinking of how that parallels Frank's story a little bit.
I beat Frank up about this every chance I get because I begged him for, like, ten years to come over to data and specifically analytics and business intelligence because Frank is a gifted natural artist. He's one of those people that can draw. And I'm almost 60 years old. I still can't color in the lines. So I had to do something like data engineering that didn't require that artistic bend. But I was thinking of that, as you mentioned, that could I use this to beat Frank up and see, I did
it's in love. Frank, you know that. Oh, I totally know. I totally know. Yeah. It only took the collapse of Silverlight and Windows Phone for me to see the light. I'm so sorry that happened. That's okay. Our second question. Lauren, what's your favorite part of your current gig? My favorite part of my current gig is talking to users of a particular product. And when the light bulb goes off between what they're saying is a pain point and a possible solution that we can build or
design, that gets really exciting to me. And so you can get a little overwhelmed by all of the user interviews that you do, especially in the beginning when you're taking in a lot of information. But then as you zoom back and then start looking at the big picture to see how you might solve some of those challenges with technology, that's where I see the real clear overlap between those user interviews and what is designed and put out into the world through tech. And
that's really exciting to me. Got you. Our next complete the sentences when I'm not working. Well, we have three questions sorry, too much coffee. We have three questions that are complete the sentence. Right. So the first one is, when I'm not working, I enjoy blank. I enjoy traveling. I love to travel as much as my time and money allow. And one of the cool things about working in Tech is that
you get to attend a lot of conferences that are in really cool places. So by virtue of being in Tech, I've gotten to see a lot of new cities and even some countries in places. For instance, I'm scheduled to go to North Macedonia next month to help teach at a tech camp in Orid, North Macedonia. And I would not
be going if not for my career in Tech. But I love to explore new places, and doing that is one of the few things that actually gets me to turn my brain off, and that's one of the things that I value about it. So I do that as much as time and money allow. I am with you. Yes. I like to not look at a calendar. That's kind of my thing. Yeah. And it's a luxury in this day and age, and when I get to do it, that's really special Macedonia. I've never been into that part of the world and I am jealous.
Yes, I'm looking forward to it. Other than Croatia, I haven't been to the Balkans. I've seen very little of Central and Eastern Europe as a region. And that's the thing about travel. As much as you've seen, there's always more to see and you know that you can't possibly scratch the surface of all of it. So I really value every opportunity that I get to see something new. Excellent. So our second complete the sentence is I think the coolest thing in technology today is blank.
¶ Technology saves time, allowing for more productivity.
I think the coolest thing in technology today is the opportunity to get time back to plan more effectively. And so that might sound like a catch 22, but I think when we look for opportunities to automate really repetitive tasks that take people hours, if not days to complete, it does give you a lot of time back to be more strategic about how you complete
the essence of your work. And so one example of that is I teach a course on interaction design at George Washington University and I had a student this past semester ask me about the impact that I think AI will have on the design profession. And I said, well, you're already using AI and design today because it's embedded into Canva and mural and all of the
software that you use to make these designs. And you're already pretty adept at using AI, but what it can't do is teach you to get really granular about the best way to design that technology to do a particular task that can solve a user need. And so I think that that is what's really cool. I think that is what is not easy to be easily automated.
And I think that if we can use technology to do the dull stuff, for instance, using natural language processing to comb through hundreds of documents and get you the information you need within minutes, that is on the surface kind of boring, but it's also hugely valuable. It's better in many cases than what humans can do and it gives you more time back. Good answer. Oh, you're on mute, Frank. Frank, I'm on mute. Sorry,
but I was coughing. The third and final complete the sentence is I look forward to the day when I can use technology to blank to drive. I would really love. I grew up learning to drive in the suburbs of Boston and then I moved to Washington DC. Which means that driving is not a fun experience for me. And I do look forward to the day when the technology for self driving cars is advanced enough that I can use it to just get in the
car, have it drive for me. I do not know what exactly that looks like beyond this idea that I just shared because obviously self Driving Cars and Regulation is a whole other podcast. But I do look forward to the day when, like, planes being effectively flown on autopilot today. I do look forward to the day when we can actually do that with cars. I wholeheartedly agree on that one. Driving in there's something about driving in and around
DC that is just an unpleasant experience. It is. And it's gotten worse over the pandemic, for sure. I notice a lot more speeding, a lot more people running red lights, a lot more people going through intersections. And as someone who straddled the border of DC and Maryland for seven years, maryland drivers are truly terrifying. And so I hope that self driving cars can alleviate a lot of that. As a Maryland resident, I do not disagree. I was
just going to interject that here in Farmville, Virginia. It's tough, too. I mean, just the other day there were like five cars at the light. It's a rough one. The struggle is real, by the way. I agree with self driving, even though it's all rural around me. Share something different about yourself, Lauren. But we remind all of our guests that we want to keep our clean rate. Yes. So something different about me is that I foster
dogs. So I have a dog myself. I have a rescue dog who is my little work from home buddy. But I also foster dogs every now and then. And so I fostered a total I did the math recently. I've fostered a total of ten within the past two years. And so every now and then I have two pups at home, and I always encourage people to foster whenever I can. We're in the summer right now.
Summer is a notoriously busy season at Shelter. So if you have ever considered fostering a dog, a cat, any other animal that just needs a home to decompress in before they get adopted, I highly recommend that people consider it. That's cool. My wife and I have done the same, and we've only managed to keep two.
¶ Adopting foster pets: fun without long-term responsibility.
Yeah, well, so one of them I did end up adopting. I did adopt one foster, but the others and people say they're like, well, is it hard to give them up? And it is to some extent, but I also think, you know, when you're a stop on their journey versus their final destination and it's hard to explain it more than that, but it is a gut feeling. And so I think you actually know, like I said,
I highly encourage people to do it. The way I also sell it to people is you get all the fun of having a pet around without the bills and long term responsibility. So that's also good if you just want a little buddy for a while but don't want a pet long term, that works out, too. It is a bit like Uber for dogs in that sense, or whatever animal. Yeah, no, we had a whole litter of puppies once that were fostered with us, and it was really cool to have that little baby puppy experience,
but. Yeah, it sounds like a lot of work, though. It was. And then as they got adopted, I was like, okay, yeah. I'm happy to see them go to their new homes where they're the center of attention. That's part of the justification for moving where we did now, where we have like, four acres, was for the dogs, basically. I work hard so my dog has a better life. Oh, totally. I work to support my dog. At the end of the day, we have a dog, but we're owned by five cats. Share it
on. That's also a good way to put it. Yeah. You're including the dog. The dog is also owned by the cats, I'm guessing. And our final question, where can people find more about you and what you're up to? Yes. So I am active on LinkedIn, so
¶ Connect on LinkedIn, visit Pragprov.com, feedback welcome.
if people want to connect to me, I would welcome that. I'm on there under my full name, and then they can also, like I mentioned, go to Pragprov.com to find the book. So that would be fantastic if your listeners want to find it and download it and then let me know what they think. So those are the main avenues. I am on Twitter as well, although less so these days. And I am trying out new platforms like Threads. I'm active on Instagram already, and so I did decide to
try out Threads as well. That is TBD, but that's used in more of a personal context. I don't talk to my friends about data governance in my everyday life, but that's also partially why I like talking to people like you about it. Cool. Well, thank you. And with that, we'll Let Bailey, our AI assistant, and the show. Thanks for joining us. Thank you, guys. Thanks for listening to data driven have you checked out Data Driven magazine yet? We are looking for
writers for the Autumn 2023 issue. Please check out Data Driven magazine.com for more information. Thanks for listening, and be sure to rate and review us on whatever podcasting app you are listening to us on.
