Turn the Lights on Data | George Firican

George Firican_mixdown.mp3-from OneDrive

George: [00:00:00] So I think everything needs to start on the business side first, so ideally, that's very clear for everybody where the business with a five year plan, if you will, for the business is so that anything else is a strategy to support that plan, right? Otherwise, it's kind of just wishful thinking. If if you want to go to Mars from a Data perspective, how can you create models for the company to be able to do that? But then if the company doesn't want to get there, then it's pointless. So that's why it's you need a business to take that first step.

Harpreet: [00:00:47] What's up, everybody, welcome to the artists of Data Science podcast, the only self-development podcast for Data scientists. You're going to learn from and be inspired by the people ideas and conversations that'll encourage creativity and innovation in yourself so that you can do the same for others. I also host open office hours you can register to attend by going to dot com forward slash a d s o h i. Look forward to seeing you all there. Let's ride this beat out into another awesome episode, and don't forget to subscribe to the show and leave a five star review.

Harpreet: [00:01:29] Our guest today is an award winning Data governance and business intelligence leader and public speaker. He's the founder of Lights on Data and its

Harpreet: [00:01:42] Youtube channel, where he

Harpreet: [00:01:44] Focuses on providing informative content such as online courses, templates,

Harpreet: [00:01:50] Guides, best practices, articles, white papers and other useful

Harpreet: [00:01:55] Resources to help you with your data governance and data management questions [00:02:00] and challenges. Today, he's here to talk to us about why data governance and stewardship, as well as master data management, is something that data scientists should care about be paying attention to and spending some time learning about. So please help me welcoming our guest today, co-host of Lights on Data show, George.

Harpreet: [00:02:23] George, thank you so

Harpreet: [00:02:24] Much for taking time out your schedule. To come on to the show, my friend. Appreciate having you here.

George: [00:02:29] Thank you so much for. It's an honor and a pleasure

Harpreet: [00:02:32] To be here with you.

Harpreet: [00:02:34] Yeah, absolutely, man. It's the first time we've connected personally one on one, so I'm excited for that. And we're involved in some group chats and I've had the privilege of having you in my happy hour,

Harpreet: [00:02:46] Been there a couple

Harpreet: [00:02:47] Of times, so it's great to finally be able to connect with you. So I'd love to learn a little bit more about you, man. So let's start by talking about where you grew up and what it was like there.

George: [00:02:57] I grew up in

Harpreet: [00:02:57] Romania, which is a country

George: [00:02:59] In Eastern Europe, and my accent, you can probably tell it's

Harpreet: [00:03:02] Not a Canadian

George: [00:03:03] Accent, but that's where I'm from. I moved out of there

Harpreet: [00:03:06] When I was about 15,

George: [00:03:08] But I had the most wonderful childhood there.

Harpreet: [00:03:10] I would say just because of the

George: [00:03:11] People close to me and family, and they really offer me an amazing life and all these experiences, which really range from, you know, working in the garden and playing with farm animals all the way from to, you know, being in the city and doing silly stuff went through a

Harpreet: [00:03:29] Revolution, the whole communist

George: [00:03:31] Regime, all that stuff. So that was an experience on its own as well. That's kind of always sticking with me, even though I was seven years old. But yeah, you know, it's all these fragments that I have in the back of my mind that really made me who I am today, I guess.

Harpreet: [00:03:45] So when you're like high school age, when so you came to Canada and

Harpreet: [00:03:50] You said, you're 15 ish?

George: [00:03:51] Yeah, I was I started grade 10 in

Harpreet: [00:03:54] Canada, so only have three

George: [00:03:56] Years before I went to university. So I spent three years in a [00:04:00] Canadian high school.

Harpreet: [00:04:01] And at that point in time, like, what did you think your future was going to look like? What what did you have your aspirations set on,

George: [00:04:08] You know, and it did

Harpreet: [00:04:08] Change from

George: [00:04:10] Year to year, depending on what I was being exposed to. But I think initially I thought it was more into design industrial design. I was playing a lot in AutoCAD,

Harpreet: [00:04:21] Which was one of my favorite courses to take.

George: [00:04:23] So that was quite, you know, revealing to me, and I thought it could put some of my artistic passion into

Harpreet: [00:04:29] Work,

George: [00:04:30] Although I realized I'm not a good drawer by hand. And even though I might be good on the digital medium, I can do almost anything by hand on paper. So I thought, OK, well, maybe that's not for me. And in the last year of university, I've kind of continued one of my passions on programing where passions at the time, because that also changed. So I thought it was going to be a program where I think by the time I finished high school, I thought I was going to do just hard, hard coding and do web development and things like that, which I did end up doing for a few years.

Harpreet: [00:05:02] But your stuff is still artistic, like you still are flexing that artistic and creative muscle with the content that you're putting out for your YouTube channel. And just like your writing and stuff

Harpreet: [00:05:11] Like the quality

Harpreet: [00:05:13] Of your videos, man is so freaking awesome. How did you get into

Harpreet: [00:05:16] Into this, this visual?

Harpreet: [00:05:19] Is it called visual design?

George: [00:05:21] Video Maybe. Yeah, maybe. You know what? Like, I always wanted to do almost everything I wanted to do to be a little bit good at everything, you know, not not master necessarily one profession. So because of that, I was trying to learn on the side all these different things, including Photoshop and Illustrator and InDesign and premiere. Yes, all of these that do have the artistic side of things, as it was something that I thought on the pastime. It's a different part of my brain that I'm flexing from what I'm learning. And then, you know, you see all these YouTubers or other people that are maybe also making an [00:06:00] income out of this. And I thought, Oh, that could be a path for me. So I started doing vlogging

Harpreet: [00:06:03] For fun and that

George: [00:06:05] Sort of branched into

Harpreet: [00:06:07] Nothing. At one point,

George: [00:06:09] But it was fun and I really learned

Harpreet: [00:06:11] Out of it, and it's

George: [00:06:13] A skill that I'm applying in my

Harpreet: [00:06:15] Data videos today.

Harpreet: [00:06:16] Yeah, I'm a huge fan. Like, that's my philosophy of learning skills is combining skills in unique ways, and I'm a fan of people

Harpreet: [00:06:23] Who are able to do that

Harpreet: [00:06:25] And execute on it in such an inventive way, right? Because it's really hard to be the best like top one. In one particular domain, let alone two or three particular domains, but you get good enough that a few different domains and you find an intersection between those. You can be the best in that intersection.

Harpreet: [00:06:43] So for you, it's

Harpreet: [00:06:44] Like, OK, you're definitely top twenty five percent when it comes to creating videos and things like that. Combine that with your expertize in data governance and now you're

Harpreet: [00:06:53] The best data

Harpreet: [00:06:54] Governance YouTuber, right? So that's that's a unique way to combine.

George: [00:06:58] Maybe the only one not too many people like to talk about this dry topic, dude.

Harpreet: [00:07:03] Only it's better than best. I guarantee you only if better than best. So, yeah, I was I was really into cad and stuff. When I was in in high school, I had a teacher, however, who made it miserable for me and I dropped that interest. It's interesting how teachers can have an effect. Like there's two classes that I really enjoyed. Yeah, but was really dissuaded because of the teachers as chemistry and cat. But I think about alternate realities. I have ended up as a chemist

George: [00:07:31] And her pretty know I was actually thinking about the same

Harpreet: [00:07:34] Thing a few

George: [00:07:35] Weeks ago because I had a most horrible experience in July high where I did have to take chemistry and we had the most awful teacher and I thought it was not

Harpreet: [00:07:43] Good at it.

George: [00:07:44] And it was such a pleasure to be in that class. And I thought, this is so complex. And then in university, I had to take chemistry one on one as part

Harpreet: [00:07:52] Of my science

George: [00:07:54] Bachelor, and I was just terrified because I knew I did badly, poorly in it. And then I [00:08:00] had the privilege of being taught by the most wonderful chemistry teacher, and I excelled in that class. And then I was thinking, like, you said, What if I had?

Harpreet: [00:08:11] I would have had much better teacher earlier on.

George: [00:08:14] Maybe I would have been a chemist. So that's true. Teachers have such a huge power and responsibility over kids.

Harpreet: [00:08:23] So what was the nudge that got you into Data, then? What was the experience that you had that that made you realize that Data was right

Harpreet: [00:08:29] For you as a great

Harpreet: [00:08:31] Teacher? Or was it just by means of working through it? How did you get involved in this Data space?

George: [00:08:36] It was working through it

Harpreet: [00:08:37] Because so I did. Early on in my career,

George: [00:08:41] I continue as a programmer and afterwards I really liked the interaction that we had the times through the business analysts with the customers

Harpreet: [00:08:51] And kind of understanding what

George: [00:08:52] Their needs are and how we are solving them, how we can program a solution for

Harpreet: [00:08:57] Them and how it's being

George: [00:08:58] Used and what the stories that come out of it. And then I started transitioning to the technical project management field, and that gave me even more exposure with seeing the impact and also better understanding the importance of Data and in particular, the quality of the data

Harpreet: [00:09:13] That it has on

George: [00:09:15] How it should be collected and how it should be maintained. And ultimately, what's the end result out of it? How do we draw information out of it through a software or visualization tool or whatnot? And so that's sort of how I transitioned from this role. That was a data quality manager role. And then I realized I kind of learned on the job that you

Harpreet: [00:09:35] Can't have

George: [00:09:36] Proper data quality without that this data governance aspect. So that propelled me to where I am today,

Harpreet: [00:09:44] You know, because

Harpreet: [00:09:45] As data scientist, machine learning practitioners, we're end users of the data, right? The day we don't get to see like the whole lineage of everything that happens

Harpreet: [00:09:54] Before all the data

Harpreet: [00:09:56] Strategy and the data management and government governance that goes into place [00:10:00] for us to be able to use that data to create whatever it is that we're doing. But I'm just curious from your perspective, what do you think the role of data

Harpreet: [00:10:07] Is in this

Harpreet: [00:10:08] New century, in the 21st century and beyond?

George: [00:10:11] Well, ideal is to empower

Harpreet: [00:10:13] People to empower people

George: [00:10:14] To make better decisions

Harpreet: [00:10:15] That are just based on facts

George: [00:10:17] And not just hunches and, you know, trends in statistics. So, yeah, ideally it's to empower people not just in your workplace, but even you as a citizen, I think we are impacted by data every day and everything that we do. And we definitely care a lot about it. When we take a look at our bank statement, we care that.

Harpreet: [00:10:40] That's correct. That's based on proper numbers

George: [00:10:43] There, right?

Harpreet: [00:10:44] And so I know you,

Harpreet: [00:10:45] You like to

Harpreet: [00:10:47] Push business leaders

Harpreet: [00:10:48] To think that

Harpreet: [00:10:50] Or adopt the mindset that

Harpreet: [00:10:52] Data actually is the asset.

Harpreet: [00:10:53] You should be valuing it as an asset, not just entirely as a cost. Yes.

Harpreet: [00:10:59] So talk to us

Harpreet: [00:10:59] About that, that kind of guiding principle you have.

George: [00:11:02] Well, so businesses have assets, that's a no brainer. The two most obvious ones out there are, well, financial assets, everything that's you're

Harpreet: [00:11:12] Touching in,

George: [00:11:14] You know, physically your computer, your your desk and so forth and so on, plus the intangible assets, but also

Harpreet: [00:11:22] Us as employees, right from

George: [00:11:24] A human resources perspective. We're also assets. And of course, Data should be an asset as well. And. We should treat it as such, so for treating four, for managing the financial assets where we have the Finance Department

Harpreet: [00:11:38] For managing people as

George: [00:11:40] An asset, we have a. So if we were to focus on the financial assets as as an example, right?

Harpreet: [00:11:46] Or who's managing it?

George: [00:11:47] Accountants, for the

Harpreet: [00:11:48] Most part, an accountant.

George: [00:11:50] Accountants are really governed by set of principles and policies, and they're being checked by

Harpreet: [00:11:55] Auditors and

George: [00:11:57] Auditors are ensuring that the correct management [00:12:00] practices of the financial assets are taking place, so there's no cooking the books.

Harpreet: [00:12:04] And what I'm seeing is that similarly,

George: [00:12:07] We should have that in place for Data. So whatever principles, policies and auditing is accomplishing for the financial assets Data governance really accomplishes for the Data side of things as well.

Harpreet: [00:12:21] So that leads me to next term. What the heck is Data governance? Anyway, I got a bunch of terms here that that I'm going to ask you about

Harpreet: [00:12:27] Because I myself,

Harpreet: [00:12:28] I'm not too familiar with them. I've only recently come across all these different terms, and I'm betting the audience probably is a little bit unfamiliar, unclear about these terms as well. So let's start with data governance. Like, what is this all about anyways?

Harpreet: [00:12:44] So Dharma is,

George: [00:12:47] You know, that's one issue of data governance. There are too many definitions for it. If you Google data governance, you'll

Harpreet: [00:12:52] Find a lot of confusion because there might be some

George: [00:12:56] Overlap, but also some contradictory definitions there.

Harpreet: [00:13:00] But Dharma is who is kind of the

George: [00:13:01] Authority when it comes to data management, and they're seeing data governance is part of data management, and they're describing it as the exercise of authority controlled and shared decision making over the management of data assets. So again, it's this body. It's just business function that creates those principles, those policies,

Harpreet: [00:13:22] Those standards that should be

George: [00:13:24] Followed organization wide.

Harpreet: [00:13:26] So like, let's kind of think about it in terms of real world, right? Let's say Data, is this generated as a byproduct of some activities, whether that's interacting through some

Harpreet: [00:13:36] System or even just,

Harpreet: [00:13:37] Let's say, coming into a doctor's office and having them write down whatever your measurements and whatnot? Mm hmm. So at what point in the process, like how does data governance impact that? Does data governance

Harpreet: [00:13:48] Say OK for you

Harpreet: [00:13:49] People who are doing data

Harpreet: [00:13:50] Entry? These are the things

Harpreet: [00:13:51] That you need to put in these fields. They could take on these values.

Harpreet: [00:13:54] If you put something

Harpreet: [00:13:55] Wrong in, we'll flag it. Is that where the governance aspect comes [00:14:00] into play, or am I kind of not grasping that correctly?

George: [00:14:03] Yeah, I think that's definitely part of it. So let's come back to

Harpreet: [00:14:07] That and think of H.R. again.

George: [00:14:09] So H.R. is the one responsible that's creating those job descriptions.

Harpreet: [00:14:13] They're creating

George: [00:14:14] The process

Harpreet: [00:14:14] For, well, here's the

George: [00:14:16] Career progression.

Harpreet: [00:14:17] Then one could take maybe start as

George: [00:14:19] A data analyst, then you become a senior data analyst. Maybe then you become a data scientist. These are the pay grades for everything. These are the responsibilities. If you want to fire somebody, this is the

Harpreet: [00:14:29] Process, right?

George: [00:14:30] So H.R. doesn't do all of that. Sometimes this is under a manager's responsibility on how to manage their own team if they want to promote somebody or fire somebody or hire somebody. They tend to do that themselves. Sometimes they're being helped by

Harpreet: [00:14:46] H.r., but whatever they do, they're really doing it

George: [00:14:49] Based

Harpreet: [00:14:50] On all these HR

George: [00:14:51] Processes, policies and guidance that they're getting. So they're not just running amok and offering a one million salary to somebody just because they think

Harpreet: [00:15:00] They can know they have

George: [00:15:02] To abide by the rules that H.R. is putting, even though it's this managers that's acting on those roles. So back to your example, yes, Data governance

Harpreet: [00:15:11] Kind of creates all of that

George: [00:15:13] And

Harpreet: [00:15:14] Tells people, well, anybody that's doing

George: [00:15:16] Data entry, you should respect these standards. These are the standards that are taking place. Now they're also doing

Harpreet: [00:15:23] That again with the help of the Drake

George: [00:15:25] Managers of the data entry people. It's not the data governance office that's knocking on your door and keeps track of it all the time. They're putting those processes in place and the

Harpreet: [00:15:35] Framework for

George: [00:15:36] Managers to also be able to do that or what they sometimes call data stewards.

Harpreet: [00:15:41] Yeah, that's another term that quite unclear about either. So Data stewards, so can anybody be a data steward? Like, what does that mean?

George: [00:15:50] And even with this, there's different thoughts of it. So Bob Signer Robert Signer,

Harpreet: [00:15:57] Who who's quite known

George: [00:15:59] In the field of data [00:16:00] governance, Data stewardship. He says that everyone is a Data stewart because everybody's touched and impacted by data. And even though there might just be visualizing it, they're not changing it. They need to have some sort of a responsibility to flag if something is incorrect or to ask the questions or contribute to those business needs and so forth and so on. There are others on the other side who are saying, Well, no Data

Harpreet: [00:16:24] Stewardship is like a

George: [00:16:26] Particular role that you can add on as a responsibility on

Harpreet: [00:16:29] Top of other

George: [00:16:30] Job.

Harpreet: [00:16:31] Responsibilities, or you're

George: [00:16:32] Calling somebody on Data Street, but it's a very niche role

Harpreet: [00:16:37] That's well defined, and not

George: [00:16:39] Everybody will do it because you're not going to expect for the VP to be a Data Stewart. Right? They might know of its

Harpreet: [00:16:46] Importance, but you're

George: [00:16:47] Not going to expect them

Harpreet: [00:16:48] To go

George: [00:16:49] And do data entry

Harpreet: [00:16:50] Or raise an issue

George: [00:16:52] And things like that. So it depends depending on the philosophy and whatever you choose to adopt their companies, for example. I think Udemy is one of the things that. Or it treats everybody as a data analyst or a data scientist. None is, even though that's not their job description, but they're offering in-house training for anybody that wants to delve further into their data. They have the opportunity to do so because they think those with the business knowledge over those areas would be able to create some insights or even come up with some questions that a data scientists otherwise in its own department would never think of it because they don't have that exposure. So again, there are companies

Harpreet: [00:17:34] That think everybody who

George: [00:17:36] Is a data scientist, even though that's not their responsibility, but they have the opportunity to be one and in similar fashion companies that think everybody is a data steward.

Harpreet: [00:17:45] Also, thank you for that. An excellent point about

Harpreet: [00:17:47] Data analysts, data scientists that you folks stick around. We're going to talk about the difference between the two because George just released a awesome video.

George: [00:17:54] Oh boy.

Harpreet: [00:17:56] Earlier today, which by the time

Harpreet: [00:17:57] You guys hear this, it'll be

Harpreet: [00:17:59] Way far in the future. [00:18:00] But I'll link to it. But but

Harpreet: [00:18:01] I digress. So just just so the data scientists and the audience can can kind of understand how data governance is

Harpreet: [00:18:08] Impacting, let's say, the day to day

Harpreet: [00:18:10] Work that they do.

Harpreet: [00:18:12] So whenever we're

Harpreet: [00:18:14] Perhaps getting access to like

Harpreet: [00:18:16] A database, right, we don't

Harpreet: [00:18:18] Have access to every single

Harpreet: [00:18:19] Database in the entire

Harpreet: [00:18:20] Company. There's certain select ones that we might have access to, but certain ones that we might have to send a ticket to it. Then it's to check whether we have the right clearance and then have to get the escalated the managers and so forth with those kind of things be what data governance touches on.

George: [00:18:38] Yeah. So in that particular example, Data governance would work very closely with like the privacy office or whatever you're calling

Harpreet: [00:18:46] It, privacy and security

George: [00:18:47] Or something of the sort. But data governance works behind the scenes a little bit before that, finding a way to classify that data to know should should we consider giving her corporate access

Harpreet: [00:18:59] To this one?

George: [00:19:00] Or it's maybe above his pay grade or it's not in his job function to actually be able to see, like the salary data that he's asking for? Yeah. So that's one of the things that data governance trying to do behind the scenes before, I guess the security team or the team is able to grant your request.

Harpreet: [00:19:18] I think that kind of helps helps you get

Harpreet: [00:19:21] A better understanding of kind of where data governance like impacts us, how it fits into the pipeline and everything. So thank you for that. So there's a couple of the terms that I'm really not familiar

Harpreet: [00:19:30] With that I'm hoping you can help

Harpreet: [00:19:31] Help me clear up on.

Harpreet: [00:19:33] Yeah.

Harpreet: [00:19:33] Metadata M.

Harpreet: [00:19:35] Data Like,

Harpreet: [00:19:36] What? What do these have to do with data? First of all, what are these things? Why do they sound so similar

Harpreet: [00:19:42] And

Harpreet: [00:19:42] What do they have to do with data governance?

Harpreet: [00:19:44] Yeah.

George: [00:19:45] Well, let's start with metadata. So metadata is described, and I don't like this definition, but it's described as data about data that doesn't tell us too much. But the way I like to think of it is through this example. So let's say you have the value. 10 That's a piece [00:20:00] of data,

Harpreet: [00:20:00] Though it doesn't tell you much if we say

George: [00:20:03] 10 years or that's a piece of metadata already years because it's

Harpreet: [00:20:07] Putting the

George: [00:20:09] Number 10 in context.

Harpreet: [00:20:10] But even

George: [00:20:11] That that's not enough, you know, is

Harpreet: [00:20:13] It years of a

George: [00:20:13] Product? Is it the

Harpreet: [00:20:15] Years of a person,

George: [00:20:16] Years of a company? So all of that information kind of adds on

Harpreet: [00:20:20] And gives us what's called

George: [00:20:22] Business

Harpreet: [00:20:22] Metadata.

George: [00:20:23] There's also a technical metadata which like for the number 10, it will tell us that it lives in this table for this column. It's flagged as an integer. It shouldn't be a double, you know, things like that. So that's the technical metadata stuff, and this information is really

Harpreet: [00:20:40] What data governance

George: [00:20:41] Is trying to govern is one of the

Harpreet: [00:20:44] Things is trying

George: [00:20:45] To find consensus into these definitions. Who should be defining it, who should own it, who's responsible for it, and then what are the data quality standards where the rules processes for pulling this information, surfacing it and again, providing that proper context?

Harpreet: [00:21:03] Speaking of understanding shared terminology and things like that, like, you've got an awesome course that's up. That's a business glossary course, right? Yeah, that would kind of be helpful. First time I ever heard of the term business, glossary

Harpreet: [00:21:16] Was, you know, as I started

Harpreet: [00:21:18] Embarking on this Data strategy journey that somehow found myself leading the charge for and yet so business. Glossary It's it's you've got an awesome course on it. You've got an awesome write up on it as well. I'll be sure to link to that. In the show notes,

George: [00:21:32] I appreciate it, yes, and again, that that covers that business metadata that I was mentioning.

Harpreet: [00:21:37] Ok. And so then what's this a master Data?

Harpreet: [00:21:40] Yeah. Master Data

Harpreet: [00:21:41] Versus metadata. What's the?

George: [00:21:43] So master data is what's called Data

Harpreet: [00:21:46] About the business

George: [00:21:47] Entities

Harpreet: [00:21:48] That provide context

George: [00:21:49] For business transactions.

Harpreet: [00:21:51] So let's think

George: [00:21:52] Of them as the most commonly found categories in your business. So things like customer, [00:22:00] that's the most obvious one employees, suppliers, products, anything that we can tie in transactional data to. It's things that don't usually change as much. It's it tends to be non transactional in nature.

Harpreet: [00:22:18] And so these types of Data like why should Data scientists

Harpreet: [00:22:22] Care about this?

George: [00:22:23] Well, so with metadata, you need to care about it to again get the context and understand it and know how should you work with it in your algorithms if you need to, especially if you need to do any transformations to it? And again, getting the context with the master Data. The issue is an organization tends to have multiple systems that stores the similar version of the Data, where the same

Harpreet: [00:22:48] Data about the same

George: [00:22:50] Master data. So you basically have George Freakin' as the customer. That's a master Data instance, but you can find some information about George in the H.R. system and the I don't know online shopping system in your CRM and maybe some other spreadsheets as well. And as a data scientist, usually you want to get the full picture in order to estimate something both this this George person. And so you need to you need to make sure that if there's ever contradictory information about George, maybe the age might be

Harpreet: [00:23:26] Different in two

George: [00:23:28] Systems you need to understand well, which one should you trust and what do you do in such cases where it's contradicting each other?

Harpreet: [00:23:35] Awesome, man. Thank you so much for that. That really helps me understand the difference between these terms that give me a better picture as to as to what they mean and how they fit

Harpreet: [00:23:43] Into this the

Harpreet: [00:23:44] Lifecycle of Data. So let's get more into kind of Data governance in action in the workplace. Starting off with

Harpreet: [00:23:53] Whether it's ever too late

Harpreet: [00:23:56] For an organization to start implementing a Data governance [00:24:00] program?

George: [00:24:01] Definitely not. No, I think it's instrumental for various reasons, but I think it's really bringing that

Harpreet: [00:24:09] Clarity over

George: [00:24:10] In so many areas on how Data should be managed as an asset. So, no, I don't think it's ever too late. I don't know if it's ever too early, either. And when you look at data governance in certain ways we

Harpreet: [00:24:25] All practice,

George: [00:24:26] Parts of it is just not really formalized.

Harpreet: [00:24:31] Yeah, yeah. I mean, I'm at a company that's been around for well over like 70 years and really haven't implemented a Data strategy and really going through a digital transformation. So that's a positive sign, but no Data governance in place, and it's really difficult to try to wrap my head around. Ok, man, how the heck do

Harpreet: [00:24:51] I even start? Like, first of

Harpreet: [00:24:52] All, I don't know much about data governance and how to make it part of my role. I don't even know how to start, so maybe you can help us understand how you would go about designing a Data governance program for organizations

Harpreet: [00:25:07] Assuming that our organization

Harpreet: [00:25:09] Lacks one and how you'd see that play out over 30, 60 90 days?

George: [00:25:15] Right? Well, first, I think you need to understand the driver. Why do we why do you need to have Data governance in place and the reason why you need to understand that driver? What's the motivation for investing resources into it?

Harpreet: [00:25:28] Yes, we need it.

George: [00:25:29] But Data governance can cover so many different areas, and you have so many different data sets and data stores and different types of master data that you can really focus on everything at once. So you need to understand the driver in order to see. Where should you prioritize your efforts and why? I'm saying this

Harpreet: [00:25:46] Is because for a lot of

George: [00:25:47] Companies, for example, years ago when GDPR came into place, that was the driver. It was some sort of a regulatory compliance.

Harpreet: [00:25:55] They were going to get in trouble

George: [00:25:57] If they didn't have [00:26:00] policies, processes

Harpreet: [00:26:01] In place, a clear

George: [00:26:02] Understanding where their

Harpreet: [00:26:03] Data leaves lives, who's the owner if they need to

George: [00:26:07] Retrieve it or delete it, what is the process in doing so? So that was the motivator. And for a lot of companies,

Harpreet: [00:26:14] This is really one

George: [00:26:15] Of the big motivators. And because that's the motivator, then that's where they're investing their efforts into. For others, there's some sort of, I don't know, business imperative. I kind of use this as an umbrella for a few drivers. So some. As we see it, as you know, we as a company, we want to invest in in business intelligence or we

Harpreet: [00:26:36] Want to do data science

George: [00:26:38] Or start harnessing the power of big data. So kind of vague demands like this, or it's something a little bit more business centric, such as we want to improve the overall efficiency of our organization or improve the customer satisfaction. So kind of data science, data analytics questions, right? And but the idea of why I'm including this under this one driver as a business imperative is because they all go under the idea of knowing what are those best

Harpreet: [00:27:05] Decisions to make based on data.

George: [00:27:08] Now that being said, maybe the driver that kind of ties it all together, it's improving the quality of the data, and nobody's kind of asking for that. But you do need to achieve that

Harpreet: [00:27:18] First if you want to do

George: [00:27:19] Everything else, even if you want to achieve that regulatory compliance. Does that make sense so far?

Harpreet: [00:27:25] Yeah, so far, so good. We get some clarity around the term. When you say business driver, what does that mean? Is that like a business driver is just something like this is a proposed benefit that will get from doing this thing.

George: [00:27:38] Yeah. So it's it's one business driver could be. Listen, we need to cut down costs

Harpreet: [00:27:45] Because this and that because we lost

George: [00:27:47] Revenue due

Harpreet: [00:27:48] To COVID and we need to kind cut some

George: [00:27:51] Costs. But we need to understand where we might need to

Harpreet: [00:27:54] Cut some costs

George: [00:27:55] From our supply chain management

Harpreet: [00:27:58] Process or [00:28:00] we

George: [00:28:00] Want to

Harpreet: [00:28:01] Attract more

George: [00:28:01] Customers or we're seeing that we're getting that

Harpreet: [00:28:04] Feedback on our

George: [00:28:06] Some of our products or the customer experience. So we want to improve that. How do we go about doing it? How do we improve our market reach over in the U.S.? Because that's where we're lacking for whatever reason. So we need the help of the Data to see where would we make the biggest impact?

Harpreet: [00:28:26] So in these roles, let's say when we're trying to implement a data governance strategy

Harpreet: [00:28:33] And

Harpreet: [00:28:34] When it comes to business

Harpreet: [00:28:35] Drivers, they should

Harpreet: [00:28:36] Should they come from like the executive team, like they are they the ones that kind of mandate the business driver? Or is it up to the data governance

Harpreet: [00:28:44] Folks who are

Harpreet: [00:28:45] Working on this initiative to maybe talk to a bunch of people and then figure out what is common among them that kind of rises up to the top and then use that as drivers like? I guess my question is how do we identify a driver?

George: [00:29:01] So it really does come from the top. Ideally in Data governance is going to support and the Data strategy.

Harpreet: [00:29:07] Overall, it will support those business needs. So what is

George: [00:29:10] The goal of the

Harpreet: [00:29:11] Organization at a high level? But then for

George: [00:29:15] This this year or next year, what are we trying to achieve there? And oftentimes businesses have, you know, some clear metrics in place. Hey, we want to increase revenue year after year by two percent, which

Harpreet: [00:29:27] Is nuts because you

George: [00:29:29] Won't. It's impossible to always keep on increasing. But that's another conversation, right? They have these targets that they need to achieve in the Data strategy. Part of it, I think, needs to figure out how can we support you to do that with the power of Data?

Harpreet: [00:29:46] Ok. Ok, so so if you had like, let's just say, a shot to speak with like a CEO of an organization and you wanted to figure out what the CEO's concerns were around Data governance, [00:30:00] would you just like say, Hey, what are your concerns around data governance or is there kind of a a business way to ask the question so that we can translate it into our own lingo?

George: [00:30:11] Yeah. No, I don't think I would ask them about data governance or Data, call it or anything. It's called Taylor really mentions that a lot, as

Harpreet: [00:30:18] Well as that no

George: [00:30:19] Execs don't care about the quality of the data. I mean, they care, but they don't.

Harpreet: [00:30:23] They don't want to

George: [00:30:23] Hear it the way they they care about it. So I think you need to find out what their pain

Harpreet: [00:30:28] Points are and

George: [00:30:29] Ask them, you know, what are you struggling with on a daily basis? Maybe. And are there might be an exact I would say, you know what? Every time I'm asking for this report, I'm getting sort of a different answer,

Harpreet: [00:30:42] Or

George: [00:30:43] I'm asking three different departments about how many new customers we have, and I'm getting all these different totals. It might not differ by much, but it's something that I'm struggling with and there's reasons why. But that's one issue that they're getting, or why can I just have these on demand? I just want to have the report and or see the dashboard whenever I want to. Why is it so painful to to get that, you know, in a timely fashion?

Harpreet: [00:31:10] Yeah. Matt, thanks so much for that. I appreciate that because, you know, as a data scientist myself, really just stat over glorified statistician like the last ten years of my life has been primarily just doing what felt like homework to me, like it was just grad school like, you know what I mean? So having to think in these business terms, it's been. A shift in mindset for me and having people like yourself and Scott Taylor

Harpreet: [00:31:36] To to

Harpreet: [00:31:37] Help educate us in very tremendous in my in my learning journey. So so thank you for that and thank you for answering my dumb questions, but I appreciate it.

George: [00:31:46] They're not dumb, but you know any time and I'm glad that I can. I can help. But I also wanted to just wrap up the first question that he had in terms of what are those steps? Because yes, you do need to get the drivers to understand where should you first invest your resources [00:32:00] in? But the 30, 60, 90 day is to me, the first three days again

Harpreet: [00:32:05] Is that understanding

George: [00:32:06] Peace and assessing the environment, especially when you're coming into a new role or if you're coming into a new company. I find we all come in with our own assumptions based on our previous experience and so forth

Harpreet: [00:32:18] And so on, which is not always

George: [00:32:20] Valid, and I think it's hindering if you're not investigating further. So first, you need to understand the entire

Harpreet: [00:32:27] Environment, you know, business

George: [00:32:28] And technical as well in the next

Harpreet: [00:32:30] Three days.

George: [00:32:31] That's when you should start putting that.

Harpreet: [00:32:34] This data governance framework

George: [00:32:35] Together, you know, start creating that council, which is this high, the highest body governing body in Data governance, if

Harpreet: [00:32:42] You will. You have all these business

George: [00:32:44] Representatives that are

Harpreet: [00:32:45] Basically deciding

George: [00:32:46] On behalf of the business. This is where we should focus our efforts on.

Harpreet: [00:32:51] Yeah. And then

George: [00:32:52] Find out what the scope of

Harpreet: [00:32:53] The Data governance

George: [00:32:54] Program should be.

Harpreet: [00:32:55] The first set of

George: [00:32:56] Priorities, you should start creating the domain model and let me know if you need to go into that.

Harpreet: [00:33:02] Yeah. So there's Data Governance Council now. Anytime somebody mentions council, I just immediately think of like

Harpreet: [00:33:08] It would Yoda.

Harpreet: [00:33:09] And yet I council.

Harpreet: [00:33:11] So like this council,

Harpreet: [00:33:13] Ideally, what would be the types of individuals that we would want to see on the council?

George: [00:33:20] So ideally, you would have executives.

Harpreet: [00:33:23] Ideally you would

George: [00:33:24] Have the highest representatives in your business be present in this council. And it doesn't always happen. So there's ways around it. And sometimes it's not good to have all the execs because you also need to

Harpreet: [00:33:37] Think of kind of their own

George: [00:33:39] Personality and how they might be clashing. So there's a lot of politics involved as well. That's why you also should have a very good

Harpreet: [00:33:46] Moderator, mediator, facilitator in these

George: [00:33:49] Meetings. But that's I guess another topic.

Harpreet: [00:33:52] I'm diverging there. But yeah, no, no.

Harpreet: [00:33:54] That's interesting. Yeah, because we're talking about how if we're meeting

Harpreet: [00:33:58] With execs, CEO [00:34:00]

Harpreet: [00:34:00] Level, like we don't just flat out talk about Data governance, we try to get to Data governance by asking other types of questions and assessing points. So when it is time to actually put together, like the Data Governance Council,

Harpreet: [00:34:12] Do we just say, all right, we're going to

Harpreet: [00:34:14] Where you are the Data Governance Council. We want you guys to talk to us about data governance.

Harpreet: [00:34:18] We'll be transparent like that.

Harpreet: [00:34:19] Or how do we entice them to join the council? Like, how do we tell them that this is going to be a benefit of their time?

George: [00:34:26] That's definitely a tough one. It kind of needs to be part of that business business case that you

Harpreet: [00:34:31] Need to make.

George: [00:34:33] That's why it's always helpful to have some sort of a sponsor towards this. So you're not just managing from the bottom

Harpreet: [00:34:39] Up, but you do

George: [00:34:40] Have

Harpreet: [00:34:40] Somebody as a

George: [00:34:41] Senior level. You maybe you have the president or the CEO or C-level executive that's really endorsing this whole Data governance program. And they're the ones

Harpreet: [00:34:51] That could get that

George: [00:34:52] Message across to get our executives coming from a high level individual definitely

Harpreet: [00:34:57] Has can have a lot of weight.

Harpreet: [00:34:59] So let's say somebody has got, you know, some there's a brave data scientist out there, let's just call him are pretty just because fictional character who, who, who has the backing of, let's say, the CIO. What are the biggest challenges you foresee her facing when he's starting out a Data strategy

Harpreet: [00:35:18] At this massive

Harpreet: [00:35:19] Organization?

George: [00:35:20] Oh, plenty. I mean, it's first of all, I think we are blinded by the entire ocean piece.

Harpreet: [00:35:28] There are a lot

George: [00:35:29] Of areas to tackle, and how do we just choose one cup out of it to to boil it once? So I think that's one. And that's where again, some some guidance from somebody that sits at that strategic level would help to give this individual some insight as to where should we focus on? Because whatever you're presenting to the council, you can give them a blank slate. Either you've got to come well prepared and give them

Harpreet: [00:35:55] Options, as well as

George: [00:35:57] You know your opinion as to [00:36:00] where or how should they vote on things. You know, in the end, they're kind of just there to make the decisions, not necessarily create the decisions or the options, if you will. So all the work is kind of done for them and they just maybe read over it, understand the impact, the repercussions

Harpreet: [00:36:19] Of choosing

George: [00:36:20] One option over the other, and then they're just there to give the stamp of approval.

Harpreet: [00:36:25] It sounds like I am in for a a long bit of time where I'll be. Having nightmares, I guess,

George: [00:36:33] You know what? Like in so many ways is not a rewarding

Harpreet: [00:36:36] Role to be in because

George: [00:36:38] It tends to be a lot of convincing

Harpreet: [00:36:41] And

George: [00:36:41] Understanding all these pain points and everybody else's has a different communication style on how to get that message across. Some people are more visual. Some people prefer some sort of a personal story that they could relate to. So it's it helps to understand how they prefer for things to be

Harpreet: [00:37:00] To become communicated to them.

Harpreet: [00:37:02] Oh, speaking of nightmares, speaking of stories, what can Stephen King teach us about Data governance?

Harpreet: [00:37:09] Right?

George: [00:37:09] I've done a video a while ago. Yeah, and I don't remember all the different things that I was drawing a parallel

Harpreet: [00:37:16] Parallel thing on.

George: [00:37:17] But I know the first one that he was mentioning that you have three months and he was.

Harpreet: [00:37:23] He wrote this book

George: [00:37:24] On

Harpreet: [00:37:24] How to write books, basically

George: [00:37:26] How to write novels. And he was giving all these pieces of advice. And I thought, Oh, you know, some of these really apply to Data governance as well. So one of them was that the fact that you have three months, you have three months to kind of get an idea in

Harpreet: [00:37:39] There and start writing your

George: [00:37:40] Book. If you're dwelling on it a lot longer, you're just not going to get anywhere. So get give yourself three months to create a rough draft. And kind of the same

Harpreet: [00:37:52] Is said

George: [00:37:52] About Data governance. From the moment where you get that approval, you have three months to show something to produce [00:38:00] something because otherwise it's just going to be LinkedIn out and people are going to lose their patience and they're going to forget about it.

Harpreet: [00:38:07] You kind of lose that momentum.

George: [00:38:09] So ideally within those three months

Harpreet: [00:38:12] Starts something, you know, maybe put in those

George: [00:38:15] Guiding principles together or figure out who should be on the

Harpreet: [00:38:18] Council,

George: [00:38:20] Create a mission,

Harpreet: [00:38:21] Something so it touch back on that principle. As you're mentioning, that's something I'm meant to talk about earlier. So the DMA, they've got these set of principles that they use and they have principles associated with all these various different

Harpreet: [00:38:37] Slices of the,

George: [00:38:39] Let's just say, Data

Harpreet: [00:38:40] Management. Yeah, yeah.

Harpreet: [00:38:41] So what are these principles? How do we identify the principles?

George: [00:38:47] Well, so I have a

Harpreet: [00:38:48] Few of them

George: [00:38:50] In the first one, and I think the overarching one is that Data is a strategic enterprise asset and should be managed

Harpreet: [00:38:57] As such straight forward.

George: [00:38:58] I think I mentioned that a few times, but I think that should be the guiding principle that we want to treat Data as an asset, right?

Harpreet: [00:39:05] The other piece maybe comes

George: [00:39:07] From the master. Data point of view is that there's only one version of the truth for our enterprise Data,

Harpreet: [00:39:14] Which then needs

George: [00:39:15] To be actively managed. And that's trustworthy. And I think this sentence alone encompasses so many things.

Harpreet: [00:39:21] The fact that we need to have that

George: [00:39:23] Ownership, we need to have the quality

Harpreet: [00:39:24] Of it, and we need

George: [00:39:25] To have that one version of the truth. Even though we might have different versions of it, we need to choose only one of them. And there are many others, you know, the fact that it needs to you are

Harpreet: [00:39:34] Data practices

George: [00:39:35] Need to comply with

Harpreet: [00:39:36] Legal and regulatory

George: [00:39:38] Requirements and internal policies and things like

Harpreet: [00:39:41] That. And so we were talking earlier about business drivers and that we got an idea of what principles are. Does it make sense to combine these two together, like combine principles and drivers together in order to do what?

George: [00:39:56] Yes, but I think that drivers can change also from [00:40:00] year to year. And ideally, your program would be nimble enough to address them, whereas principles, I think they're they would withstand the test of time, so to speak. You know, it's kind of like high level stuff. Let's be ethical, regardless of the driver. That's the principle that we always follow. We go back to our principle

Harpreet: [00:40:20] And say, Well, we embarking on

George: [00:40:21] This project and how we're doing

Harpreet: [00:40:23] Things.

George: [00:40:23] But are we still ethical? Yes. Ok, let's move. Move on.

Harpreet: [00:40:28] Perfect. And that makes

Harpreet: [00:40:29] Complete sense, that clarifies it for me a lot, actually. So let's talk more about Data strategy mostly like like I mentioned, like I'm I am navigating the

Harpreet: [00:40:38] Labyrinth as as

Harpreet: [00:40:40] Dana likes to say. So I've been waiting through so many books here, man.

Harpreet: [00:40:43] I've got navigating Labyrinth Modern Data

Harpreet: [00:40:46] Strategy, the Data Management Tool Kit. I got Scott Taylor's book here as well. It's becoming extremely challenging for me as a data scientist who just loves writing code and building models and deploying things to production to now think about everything that happens before I get my hands on the data.

Harpreet: [00:41:04] And I think this is such an interesting

Harpreet: [00:41:07] Place to work. You know, this juncture that I'm at in my career

Harpreet: [00:41:11] Because I'm getting a little bit

Harpreet: [00:41:13] Of the clouds, a little bit of the dirt, right? I'm seeing things from both ends. So I guess what does Data strategy have to do with helping us get ahead in our Data careers?

George: [00:41:24] So at a high level, the Data strategy is really that set

Harpreet: [00:41:27] Of set of choices

George: [00:41:29] Instead of decisions that brought together, they kind of charted high level course of action to achieve our business goals,

Harpreet: [00:41:37] Right? And to

George: [00:41:39] Me, all of this really ties in with anything that has to do with Data because part of the Data strategy

Harpreet: [00:41:45] Could be, well, we need

George: [00:41:46] To grow our data

Harpreet: [00:41:47] Science team or we need our

George: [00:41:50] Science team to achieve these things or help us come up with questions and answers on how to address these business needs. And in the end, [00:42:00] it's part of a subset of that Data strategy could be, you know, listen to. We've noticed that as data scientists, you

Harpreet: [00:42:06] Guys take

George: [00:42:07] Quite a little bit of time to deliver that end result because 80 percent of your time is being spent on cleansing the data or transforming

Harpreet: [00:42:15] It before.

George: [00:42:16] So why don't we just focus our resources there and see if we could decrease that time?

Harpreet: [00:42:22] So I guess, like how can

Harpreet: [00:42:24] We help our organizations

Harpreet: [00:42:26] Define a

Harpreet: [00:42:27] Data strategy if we find ourselves in this position of having to to build a Data strategy? I would like what are some things that we should think of. Like we talked about the different terms like this data management, there's master data management, there's metadata management and data governance, I guess. Do all of these things happen concurrently all at once? Or is there like a chicken and egg problem going on here? Like, where does one start?

George: [00:42:51] So I think everything needs to start on the business side first. So ideally, that's very clear for everybody where the business with a five year plan, if you will, for the business is

Harpreet: [00:43:03] So that anything else is a strategy to support that plan,

George: [00:43:09] Right? Otherwise, it's kind of just wishful thinking. If if you want to go to Mars from a Data perspective, how can you create models for the company to be able to do that? But then if the company doesn't want to get there, then it's pointless.

Harpreet: [00:43:24] So yeah, that's

George: [00:43:25] Why it's you need a business to take that first step.

Harpreet: [00:43:29] And are there any blueprints that that kind of exist to help create a Data strategy? If so, like how do we even use the blueprints?

George: [00:43:39] Yeah, that's that's a good question. And yes, I would say that there there are in some aspects I could think of maturity models as an idea on how it could offer you a blueprint because the maturity model

Harpreet: [00:43:54] Is kind of giving you a

George: [00:43:55] Way to assess where you are,

Harpreet: [00:43:57] But not only that to let you know what [00:44:00] the next step

George: [00:44:01] In that strategy should be. And then having the delta between them to you kind of see what steps you should take.

Harpreet: [00:44:10] Speaking of maturity models that you gave me some access to your question, which I'm grateful for. I really enjoyed that course.

Harpreet: [00:44:16] And again, maturity model

Harpreet: [00:44:18] Is something I had no idea about until

Harpreet: [00:44:20] Recently, until like

Harpreet: [00:44:21] A year ago. So what? What the heck are the maturity models like? What are they all

Harpreet: [00:44:27] About and

Harpreet: [00:44:28] Why on earth are there's so many of them?

George: [00:44:31] So, yeah, so as I mentioned, it's really just a way for an organization to

Harpreet: [00:44:35] Assess their improvement

George: [00:44:37] In a particular discipline. And you can have maturity models for Data governance, for Data management in general, for even data science, things for AI. So there's all kinds of maturity models, but it really offers a company a way to to see where they are today, where they were yesterday. Is there any improvement or

Harpreet: [00:45:00] Not where

George: [00:45:01] They want to be tomorrow and how do they get there?

Harpreet: [00:45:04] I'll definitely be sure to

Harpreet: [00:45:05] Link to that course, and I think you had like a really cool blog post about it as well that I really enjoy. It's all linked to both those

Harpreet: [00:45:12] In the show notes. Thank you. And as

George: [00:45:15] To why there are so many, well, everybody has their take

Harpreet: [00:45:18] On it.

George: [00:45:18] And a lot of these maturity models are tend to be tied to the usage of software and tools. So normally you would have all these soft. Or vendors that are trying to sell you the blueprint for the Data strategy, and it's part of the blueprint as well, if you get to use this one, you were to advance. If you get to use our tools or even our consulting services

Harpreet: [00:45:41] Or whatnot, then

George: [00:45:43] So it's yeah a way for them to pitch in whatever they're selling.

Harpreet: [00:45:48] Can we have the George tech and maturity model? Does that exist?

George: [00:45:52] No, he doesn't.

Harpreet: [00:45:53] So how do we get buy-in then from from leadership to go along with this? With this harebrained idea, [00:46:00] we've got to implement a Data strategy. It's going to give them all this benefit, but I can't articulate what the benefit is that it's going to give to them or show them anything tangible. How do we how do we start about getting the buy in?

Harpreet: [00:46:12] Yeah.

George: [00:46:13] And it's definitely a tough, tough piece to calculate that return on investment. Sometimes that's what they care about. If they're financial driven person, they want to see if I'm going to invest this. What are the cost savings that I'm going to get out of it? There's different calculations that you could you could create to

Harpreet: [00:46:29] Show how

George: [00:46:31] Having Data governance in place Data quality management piece in place could save on, you know, data quality

Harpreet: [00:46:38] Errors. And again, as I

George: [00:46:39] Mentioned, there's different types where you can calculate it. If you're catching the Data quality error at

Harpreet: [00:46:44] Entry or, well, it's part of a report

George: [00:46:47] Or what happens if it's actually used in transformed into

Harpreet: [00:46:50] Information, which tends to be bad information if it's based

George: [00:46:53] On wrong data. So you could, yeah, base it on that, which is very straightforward calculation. I think the first one is like ten dollars or no, $1, $10 a hundred dollars

Harpreet: [00:47:04] As an estimate. I think it's something

George: [00:47:05] That Deloitte or IBM initially put together,

Harpreet: [00:47:08] So you can refer to that. Or you can

George: [00:47:10] Even think there's different surveys that show the amount of time spent by data scientists or just by the business information worker on how much time they spend to look for information.

Harpreet: [00:47:23] Look for Data run

George: [00:47:24] Reports so you can use that as a basis on, well, let's tackle these issues and how we can improve it or that regulatory compliance piece. I forgot which bank was just fine last November, like $400

Harpreet: [00:47:38] Million because they didn't

George: [00:47:40] Have data governance

Harpreet: [00:47:41] In place. They didn't have

George: [00:47:42] Proper privacy and security practices. That could be another good example to compare yourself to your industry peers and see where do

Harpreet: [00:47:50] You hit by fine

George: [00:47:52] And what's the cost avoidance of that? So that's kind of the stick method and the care method. Again, I think it's to see [00:48:00] what could be done with these cool AI machine learning projects that you could do. What do you need to have the data and clear understanding of it and not have to

Harpreet: [00:48:10] Keep on cleaning it every

George: [00:48:12] Time? And even that there might be data that you don't have in place. So how can we get

Harpreet: [00:48:16] There, what we need to

George: [00:48:17] Have this infrastructure in place first?

Harpreet: [00:48:20] And that's been such a wonderful conversation about data governance.

Harpreet: [00:48:24] I've definitely learned a lot, both through

Harpreet: [00:48:26] Your videos,

Harpreet: [00:48:27] Through the online course

Harpreet: [00:48:28] That gave me access to which I really enjoyed.

Harpreet: [00:48:30] I just this conversation as well.

George: [00:48:33] Let's time Harpreet Sahota.

Harpreet: [00:48:35] Oh yeah. Appreciate it, man.

Harpreet: [00:48:36] Let's transition the conversation towards the video that you recently

Harpreet: [00:48:39] Put out, which I found fascinating.

Harpreet: [00:48:41] So talk to us about talk to us about your most recent video where you talk about data analysts versus data scientists. And like all the effort that you put into creating that. Walk us through what you did there.

George: [00:48:53] Yeah. Well, so I do work with data analysts as part of my team, and they're amazing people. They're a lot smarter than I am, and I feel very privileged with everything that they're doing in teaching me. And there's always that discussion happening, you know, in live or online as to what the difference is between a data analyst and a data scientist.

Harpreet: [00:49:13] So being that

George: [00:49:14] I am in data governance, I kind of want to find and provide clarity overall. So it's starting to do some research, and I started by looking at all these different job descriptions to see what what do

Harpreet: [00:49:26] Companies think these two

George: [00:49:29] Rules are? And I was

Harpreet: [00:49:31] Confused.

George: [00:49:32] After a hundred or so job

Harpreet: [00:49:34] Descriptions, I came more

George: [00:49:36] Confused than I was in the beginning, so I thought, OK, well, that's not going to take me anywhere, though I found some overlaps, which they documented, and

Harpreet: [00:49:45] That helped me create some of the

George: [00:49:47] Content for the course. And then I also looked into a bunch of articles in white papers, especially from universities, to kind of see how how they what

Harpreet: [00:49:55] They think of it all.

George: [00:49:56] So overall, it's not 100 percent

Harpreet: [00:49:58] Clear to me yet,

George: [00:49:59] But I [00:50:00] hope I did provide some some guidance by looking at the the scope of both roles, the the skills required in both the that education and experience that's needed in India and also the salary that comes

Harpreet: [00:50:12] Tagged to to each role. So it's definitely a process.

George: [00:50:15] It took a while and then I kind of spend the weekend to film

Harpreet: [00:50:18] And edit and

George: [00:50:20] Compile, and that was my free fun time that was spent on this video.

Harpreet: [00:50:24] Yeah, it's an interesting way that you did that got a lot of.

Harpreet: [00:50:27] Effort to go in and really do this

Harpreet: [00:50:30] Compare and contrast and like, what would you say, like, like what was the biggest takeaway from undergoing

Harpreet: [00:50:36] All this effort? Like what?

Harpreet: [00:50:37] What is the difference between data scientists and data analysts?

Harpreet: [00:50:40] Is it? I know you said you're

Harpreet: [00:50:42] More confused coming out than you were going in,

Harpreet: [00:50:44] But was it any

Harpreet: [00:50:45] Thing that can definitively make one analyst or a scientist?

Harpreet: [00:50:51] Well, in the end, I think I did

George: [00:50:54] Find a difference, but again, I think it's also reflecting my own opinion. So and that's something that I'm standing in the video. The fact that if you disagree, please let me know. And what is your take on it?

Harpreet: [00:51:05] Some people identify

George: [00:51:07] One or the other, or they identify themselves as a data analyst, though from other people's perspective, they might have data scientists responsibilities in there. So to me, the data analyst is really focusing more at a micro-level on on a particular question, such as how do we increase revenue in this area

Harpreet: [00:51:27] Or why did our

George: [00:51:28] Marketing effort went better in this area versus the. And they're mostly answering that question through descriptive analytics and looking at some sort of a structured data set,

Harpreet: [00:51:41] Whereas the

George: [00:51:42] Data scientist, I found

Harpreet: [00:51:44] That they're more at

George: [00:51:45] This macro level working with structured and unstructured data and not necessarily answering this

Harpreet: [00:51:51] Question.

George: [00:51:52] But maybe coming up with this question to. So let me give you an example. Actually, we'll go back to H.R. as well. H.r. is maybe [00:52:00] coming to the data science team saying, You know what? We need some help to figure out how could we reduce the spending on recruitment? Help us figure it out. Look at the data, find an answer. And the data scientist is starting to

Harpreet: [00:52:13] Look at, you know, all

George: [00:52:14] These different data sets to

Harpreet: [00:52:16] Figure out what could they

George: [00:52:17] Do? What is the hypothesis there that they need to formulate? And they figure out

Harpreet: [00:52:22] Actually out of the

George: [00:52:23] Data that they have access to that there's a high turnover rate in one department. So maybe now the effort should not be on how to reduce the recruitment cost. But can we create some sort of a model where we could identify those people that are on the verge of leaving? Or they're more likely to leave the company so we could identify

Harpreet: [00:52:46] Them early as a potential

George: [00:52:48] Risk

Harpreet: [00:52:48] Factor flight risk

George: [00:52:50] And address that in a timely manner? Right. So they're coming up with the question as

Harpreet: [00:52:55] Well

George: [00:52:57] As a solution afterwards, if that makes sense.

Harpreet: [00:52:59] Yeah, definitely. Like, I've got a

Harpreet: [00:53:01] Pretty liberal view about data science to me, like business intelligence. You know, certain business analysts, data analysts, data scientists, machine learning. We're all data scientists. At the end of the day, we all fit in this umbrella of data science. I mean, definitely, I mean, I I definitely feel like data governance, data management and data management, that's all part

Harpreet: [00:53:22] Of the data science family of things.

Harpreet: [00:53:24] I don't know. That's just my

George: [00:53:26] It's part of the same ecosystem. Yeah.

Harpreet: [00:53:29] So you're talking about like structured data, unstructured

Harpreet: [00:53:32] Data, dumb question here.

Harpreet: [00:53:33] Does data governance care about unstructured data or is it only about structured data like how's that?

George: [00:53:38] Most often it's about structured data, or that's the initial focus. And now, you know, again, we're thinking about different definitions, but with unstructured data that covers like documents and PowerPoint presentations, emails, things like videos that could be seen as part of the information governance, you [00:54:00] know, and how do you structure your files and how do you what's the folder structure and how do you assign permissions to those and that whole records management piece? And sometimes it's part of the same like information governance, data governance.

Harpreet: [00:54:14] They're part of the same portfolio.

George: [00:54:16] Sometimes they're they're different.

Harpreet: [00:54:18] There's levels to this

Harpreet: [00:54:19] Stuff, man, that

Harpreet: [00:54:21] Definitely levels to this stuff. So, George, last final question before we get to a random round. Ok, so the last formal question, rather not final question. There's still a bunch of questions, so it's 100 years in the future. What do you want to be remembered for? Wow.

Harpreet: [00:54:38] And I don't quite know. Maybe, maybe

George: [00:54:40] This video that I

Harpreet: [00:54:41] Created today or

George: [00:54:43] Maybe this interview.

Harpreet: [00:54:45] Hey, man, that'd be awesome. I'm sure you'll be remembered for for doing far greater things than being on this interview, but I appreciate that, man. So let's jump into this man. So you are a YouTuber yourself creating awesome content, awesome videos. When do you think the first video to hit one billion views on YouTube will happen, and what will it be about?

George: [00:55:07] I'm thinking maybe in the next five years

Harpreet: [00:55:10] I forgot what the highest one is,

Harpreet: [00:55:12] But it's a few beats

Harpreet: [00:55:14] Baby shark. Yeah.

George: [00:55:16] And you know what? I think in the top 10, there's actually a couple other

Harpreet: [00:55:21] Nursery

George: [00:55:21] Rhymes in there

Harpreet: [00:55:23] And at least songs that

George: [00:55:24] Cater to kids, right? So I'm thinking. It would be something that's appealing to kids, as at first trillion views YouTube VIDEO Yeah.

Harpreet: [00:55:33] And when do you think that would be said five years? Ok, so five years? Yeah. All right. So I've been collecting data on this. I'm going to publish publish this. Research in the future will be groundbreaking. So maybe you'll be remembered for the guy who got the prediction about the one trillion views actually right?

Harpreet: [00:55:54] So in your

Harpreet: [00:55:54] Opinion, what do most people think within the first few seconds of meeting you for the first time? [00:56:00]

George: [00:56:00] I think they're thinking, why do I have such unruly hair? I think the first impact is sort of visual, you know, and that's sort of the visual aspect. My hair

Harpreet: [00:56:09] Is out of

George: [00:56:10] Control most of the time.

Harpreet: [00:56:12] You have long hair, right? I used to to work. Yeah, right. Yeah, yeah, yeah.

George: [00:56:17] And my wife really prefers longer hair. So we always have this argument

Harpreet: [00:56:20] And how

George: [00:56:21] You know how much I could have the haircut?

Harpreet: [00:56:24] Well, yeah. Well, speaking of looks like who do people tell you that you look like it,

George: [00:56:30] Jude Law or Johnny

Harpreet: [00:56:33] Depp? And both don't make any sense.

Harpreet: [00:56:35] So yeah, I'd say more like Leonardo DiCaprio, man. Yes, yes.

George: [00:56:40] Seriously, especially since he got a little bit chubby, so I can do that.

Harpreet: [00:56:46] But what are you currently reading?

George: [00:56:48] You know what? I have it right here. I'm reading the series

Harpreet: [00:56:52] From TED

George: [00:56:53] Ted.com. So out of all their their speeches and presentations, they they actually released all these books, and

Harpreet: [00:57:02] There's one that I'm reading

George: [00:57:03] Right now that says how we'll live on Mars. So it's a series of three basically presentations in written format that address these topics.

Harpreet: [00:57:13] This is

Harpreet: [00:57:15] Ellen in any of those

George: [00:57:17] Conversations there. Ellen actually is not in here. No.

Harpreet: [00:57:21] What song do you have on repeat?

Harpreet: [00:57:23] Oh boy, you know what? I need to take a look

George: [00:57:26] At the the

Harpreet: [00:57:27] Name of it because I'm not

George: [00:57:29] Very good with names. Believe it or not. And one second. So there's a song that I discovered it's from the 70s, believe it or not.

Harpreet: [00:57:41] And it's from there you go. It's called live for today.

George: [00:57:45] Ok, who's up by the grass grass roots? All right. That's a check. And yeah, I listen to it a couple of weeks ago and I thought it was so catchy and so positive. And you have been playing it ever since. Every time I go out on a walk or I'm [00:58:00] feeling down like a put it on.

Harpreet: [00:58:02] Definitely check that out right after this. So I'm going to open up

Harpreet: [00:58:05] The random

Harpreet: [00:58:07] Question generator. So we rule fun here. And the first question out of the random question generator is pet peeves, OK?

Harpreet: [00:58:15] You know, when

George: [00:58:16] You're using the microwave and maybe you're stopping the microwave before the timer goes out? Mm hmm. I'm really upset if if you don't restart that timer, you know,

Harpreet: [00:58:28] And my wife does does this

George: [00:58:30] A lot and then I don't see whatever the the time

Harpreet: [00:58:33] Is on the

George: [00:58:34] Microwave or

Harpreet: [00:58:35] Clock there. But I feel funny, man.

Harpreet: [00:58:37] I also try to avoid the final microwave beep. Yeah, but I always reset it. And I was like, What's on your bucket list this year?

George: [00:58:47] A few things. One is a surfing trip, which I try and do every year, and ideally I would love to go travel outside of the country, go someplace where I'm really craving for that. Thailand, Mexico, somewhere on a beach.

Harpreet: [00:59:04] So, so you're a surfer, huh? So how long have you been surfing for a

George: [00:59:07] Few, maybe seven years now, but I'm really

Harpreet: [00:59:10] Bad at it, though I would

George: [00:59:11] Imagine you're from

Harpreet: [00:59:12] California, right? I am.

Harpreet: [00:59:14] Yeah, I've never I don't even know how to swim, man. I've never surfed and I've never, never gone for. I mean, I've swam, but I'm not really good at it. But that's crazy. My seven years I've been surfing like wears some crazy places. You've been surfing at

George: [00:59:27] Surfing maybe once or twice a year. So again, maybe that's

Harpreet: [00:59:30] Why I'm not good.

George: [00:59:31] We usually go here on the West Coast, on Vancouver Island, in this area called Tofino. It's just amazing. It's beautiful. It's one of my favorite places on the West Coast, so mainly here. But also we visited our Californian friends a few times,

Harpreet: [00:59:45] So we've tried things

George: [00:59:47] There. But it feels overwhelming because there are so many professional surfers in the waters and you just feel inadequate.

Harpreet: [00:59:53] At least you're having fun, man. That's the most important part, right? Yeah, yeah. So you know, it's been on my list to go to.

George: [00:59:59] So [01:00:00] that's right. Let me know.

Harpreet: [01:00:01] Let me know.

George: [01:00:02] I'll I'll give you a private tour.

Harpreet: [01:00:05] Nice, man.

George: [01:00:06] Let me know so that and I also want to finish two courses like create two courses on Data governance.

Harpreet: [01:00:11] So, yeah, you've created quite quite a few courses, so you got the course

Harpreet: [01:00:14] On the business.

Harpreet: [01:00:15] Glossary Like we're talking about just now and the course that I was just talking about the maturity models, what other

Harpreet: [01:00:20] Courses do you offer?

George: [01:00:22] I have another one with a dear friend and colleague, Annabel Santos, which is on Data visual. For Data storytelling and is just really covering best practices on that and how you can tell the stories with Data visualizations.

Harpreet: [01:00:34] Very nice. Do you ever sing when you're alone? What songs

George: [01:00:39] I do? I don't want to say what songs I do. I'm a very bad singer, and this is maybe one of the things I wish I. I had like a singing voice, so that's why I do lip synching sometimes. But I think you're you're you've been challenged by Susan Walsh, actually.

Harpreet: [01:00:54] Have I got to take it up on that, dude? I'll do it. Yeah. Challenge. Yeah, yeah, you have. Or, you know, take her up on a lip singing challenge as well.

George: [01:01:00] So I challenged her and she challenged you. And then you. You feel like you drop the ball there. So, but maybe you've never seen

Harpreet: [01:01:07] The comment I got.

Harpreet: [01:01:09] I got to find that common, man. I will not back down. Next question here the when people come to you for help. What do they usually want help with?

George: [01:01:18] It's usually something technical

Harpreet: [01:01:22] Or something to like my family members,

George: [01:01:24] For example, they ask like, Oh, can you enhance this photograph for me or can you carry this like me? Video for B or show me how to do something with my website or with my social media?

Harpreet: [01:01:36] So, yeah,

George: [01:01:37] Little things like that.

Harpreet: [01:01:38] Let's do one more from here.

George: [01:01:39] Favorite city will be Tofino. I think I'm really in love with a couple.

Harpreet: [01:01:46] Couple of places, actually.

George: [01:01:47] One is Bali as a region, been there only once, but I just felt so serene and I really want to go back. Just felt at peace over there. But I really enjoyed this. This European city in Italy called [01:02:00] Florence, had been there a couple of times and again, it's just something about the the vibe in the sunset. And yeah, the whole feel

Harpreet: [01:02:06] Of the city might not

George: [01:02:08] Be the prettiest out of Italy, but it just kind of spoke to me.

Harpreet: [01:02:12] That's awesome, man. So, George, how can people connect with you, what can they find you online,

George: [01:02:17] Just LinkedIn or lights on Datacom?

Harpreet: [01:02:20] George, thank you so much for taking time out of schedule to be on the show today, man. Appreciate having you here.

George: [01:02:24] Thank you so much for preaching.