My Career in Data Season 4 Episode 11: Samuel Spencer, CEO and Co-Founder of Aristotle Metadata

1 - > Shannon Kempe: Hello and welcome. 2 - > My name is Shannon Kemp and I'm the Chief Digital Officer at 3 - > Dataversity, and this is my career in Data, a Dataversity 4 - > Talks podcast dedicated to learning from those who have 5 - > careers in data management to understand how they got there 6 - > and to talk with people who help make those careers a little bit 7 - > easier. 8 - > To keep up to date in the latest in data management education, go 9 - > to dataversity.net forward slash subscribe.

10 - > Hello and welcome to My Career in Data, a podcast where we 11 - > discuss with industry leaders and experts how they have built 12 - > their careers. 13 - > I'm your host, Shannon Kemp, and today we're talking to Samuel 14 - > Spencer from Aristotle Metadata. 15 - > Today we are joined by Samuel Spencer, CEO and co-founder of 16 - > Aristotle Metadata. 17 - > And normally this is where a podcast host would read a short 18 - > bio of the guest, but in this podcast, it's your bio that 19 - > we're here to talk about.

20 - > Sam, hello and welcome. 21 - > Hi, Shannon. 22 - > Nice to be here. 23 - > I'm so grateful for you being here, and I'm so excited because 24 - > we recently met at the Dataversity Data Governance and 25 - > Information Quality conference in San Diego a couple weeks ago.

26 - > Um had so much fun chatting uh and uh got to know you a little 27 - > bit and hear a little bit about your story. 28 - > So I'm excited to hear more about how you got into founding 29 - > a company. 30 - > SPEAKER_02: It's uh it's a really long journey. 31 - > It's an interesting story.

32 - > So I'm really excited to tell it to you. 33 - > Shannon Kempe: Me too. 34 - > Okay, so so let's start with where you are. 35 - > You are the CEO and co-founder of Aristotle Metadata.

36 - > So tell me, for those who may not know, what type of business 37 - > is Aristotle metadata? 38 - > SPEAKER_02: So we're a software as a services company that 39 - > provides data governance platforms that help 40 - > organizations understand what data they have, um, who's 41 - > responsible for it, and probably more importantly, why it's being 42 - > collected. 43 - > So the way that I like to explain it is it's a little bit 44 - > like JIRA for data governance. 45 - > Um, Jira isn't just for project managers, it's a platform where 46 - > project managers can kind of communicate to everybody across 47 - > the organization.

48 - > And it's a place where, you know, if you're just starting 49 - > your career, it's a it's a tool you can use to kind of learn 50 - > what project management is. 51 - > We see Aristotle as the same for data governance. 52 - > It's a place where data governance experts like 53 - > ourselves can help everybody become a data steward so that we 54 - > can focus on doing less and reviewing more, which is, I 55 - > think, one of those lessons that they teach you in management is 56 - > like, you know, management is not doing, it's it's more about 57 - > reviewing.

58 - > So that's what we do. 59 - > Shannon Kempe: Oh, that's very, very nice. 60 - > And so tell me, as a CEO and co-founder, what is it that you 61 - > do? 62 - > What's your daily, you know, in and out look like?

63 - > SPEAKER_02: Um, I do everything else. 64 - > Um, so we we um when we started the company or when you start a 65 - > company, you do everything. 66 - > Like you do everything. 67 - > So my role was I was the technical co-founder, so I was 68 - > responsible for building the product.

69 - > So I had a background in computer science. 70 - > Um, and as we've grown, you know, lots of those 71 - > responsibilities have kind of fallen away to other people. 72 - > So uh at the moment, like probably my two biggest roles 73 - > are around training. 74 - > So uh how to how to use the platform and training on you 75 - > know new new methods and new approaches in how data 76 - > governance can be done and communicated to organizations.

77 - > And recently I've been doing a lot of research because uh one 78 - > of the things they don't tell you is when you start a metadata 79 - > software company, you actually have to sell people on metadata 80 - > first. 81 - > You can't just rely on people wanting to do it. 82 - > So over time, I've actually started to see some techniques 83 - > that are really good for communicating why data 84 - > governance is actually something you want to do. 85 - > Uh, because you know, the other thing is every CEO is a 86 - > salesperson.

87 - > So, you know, how could I help people understand, you know, why 88 - > data governance is important and how to kind of communicate that 89 - > into their own organizations to invest in data governance as a 90 - > principle? 91 - > Shannon Kempe: Very, very nice. 92 - > And so and tell me, Sam, you know, so you're building a tool, 93 - > a software as a service that uh, you know, encapsulates the data 94 - > uh and works with the data, but how do you work with data in 95 - > your day-to-day job?

96 - > SPEAKER_02: Um, a big part of it is uh a lot of my job, well, 97 - > like like every executive, a lot of my job is data-driven. 98 - > Um, so I'm looking at reports on a regular basis uh around the 99 - > performance of the company, the growth of the company. 100 - > Um, and that's things like you know, marketing metrics, it's 101 - > sales metrics, it's you know, server uptime. 102 - > So, you know, a really big part of my role is just you know 103 - > understanding what's happening because you know, one of those 104 - > hard lessons is you can't you just can't be involved in 105 - > everything.

106 - > You just have to rely on good data coming in. 107 - > And that's been really educational for me because you 108 - > like you start to see why data becomes important and where 109 - > those where those challenges come. 110 - > Um, one of one of the things that I like to say is we have, 111 - > you know, we're we've got uh you know 30 people back in 112 - > Australia. 113 - > We have probably three or four different definitions of client.

114 - > And that might sound weird. 115 - > You'd say, well, you're an expert. 116 - > Shouldn't you shouldn't you have just had one definition and it 117 - > and kept that? 118 - > And the reality is that often, you know, client will will mean 119 - > different things to different people.

120 - > Like sales actually thinks about it differently to how our 121 - > operation staff do. 122 - > You know, they think about you know what it looks like for a 123 - > client from a server perspective, not a sales 124 - > perspective. 125 - > So part of my role is is really around, you know, um just 126 - > analyzing what we're doing, how we're doing, where we can 127 - > improve. 128 - > Uh, and that's that's really helped um, probably that second 129 - > bit is around a lot of that research of you know analyzing 130 - > how people talk about it and and how we can communicate why data 131 - > governance is important is is a big thing I'm doing at the 132 - > moment is is doing some surveys into how we can help you know 133 - > professionals like you and I talk about why data governance 134 - > is something you we just have to invest in.

135 - > It's such a good thing for us to do. 136 - > Shannon Kempe: I I can totally really, I mean, being uh you 137 - > know a small company ourselves, you know, we the irony of what 138 - > we do, right? 139 - > We still have to manage our own data, right? 140 - > And we did that a very similar exercise in the company to go 141 - > through and decide what is the definition of customer, right?

142 - > What is the definition of a customer for today diversity? 143 - > And we have several definitions um as well. 144 - > SPEAKER_02: Like the thing that I like to say there is like I 145 - > think um we forget to realize that data, data comes from 146 - > people doing business, like it comes from somebody doing their 147 - > job. 148 - > Like people don't go, I'm gonna create some data today.

149 - > They go, I'm gonna run a site, a sales campaign, or I'm gonna run 150 - > a marketing campaign, or you know, they don't go, oh, I'm 151 - > gonna make some uptime data. 152 - > You go, I need to deliver servers to a person. 153 - > Um, and I think for for Dataversity, you didn't go like 154 - > we're gonna get a whole bunch of data. 155 - > You said we need a conference, we're gonna have a conference, 156 - > and to run a conference, we need attendees and we need to know 157 - > about you know what food to serve them.

158 - > You you probably didn't think data. 159 - > You thought, I want to run a conference, and data is the 160 - > secondary effect that comes from us doing our jobs. 161 - > Shannon Kempe: It's very, very true. 162 - > Okay, uh, you know, normally I save this question for later in 163 - > the interview, but I'm gonna pull it up here now because it 164 - > seems very appropriate to where we're going.

165 - > Um, tell me, Sam, you know, especially since you have 166 - > founded a company around data, what is your definition of data? 167 - > SPEAKER_02: Um I think like you probably get a different 168 - > definition every time you do this, which is probably really 169 - > interesting. 170 - > But um, the way that I like to define it is um, and I've got it 171 - > on my screen because I I talk about this in training, is data 172 - > is the information we collect about standard business 173 - > activities to help us understand how the business operates.

174 - > Um, and the reason, like the the second question you probably 175 - > want to ask is why do you define it like that? 176 - > Is when we have all of this data, when we have so much data, 177 - > um, I think that definition of tying it back to business 178 - > outcomes uh is really important for helping us understand where 179 - > the scope of data stops in an organization. 180 - > Because, you know, when you uh when I get home, we've had a 181 - > really big year um to celebrate we're gonna go out to lunch, 182 - > which means we're gonna have to get some dietary requirements, 183 - > we're gonna have to get some menu orders from the team.

184 - > We don't need that. 185 - > We don't need to keep that long term. 186 - > Like it's not gonna help our business operate better to keep 187 - > everybody's food order and everybody's preferred pizza. 188 - > Um, so being able to say like really data from a from an 189 - > organizational perspective is about what the business does to 190 - > improve how the business operates is critical, I think, 191 - > to making sure that we don't take on too much.

192 - > Shannon Kempe: Indeed. 193 - > All right. 194 - > So let's let's get into uh let's back it up here a little bit and 195 - > let's talk about how you got into your current role and and 196 - > how you got into uh knowing so much about data and forming this 197 - > opinion on where um the definition. 198 - > So tell me, Sam, when you were say six years old, what was the 199 - > dream?

200 - > What did you think to yourself? 201 - > I wanna I want to grow up to be. 202 - > I mean, did you say I'm gonna grow up and be a CEO and 203 - > co-founder of a company, or did you what was the dream at six? 204 - > SPEAKER_02: Um, I I I don't know, to be honest.

205 - > I think um I knew I was always gonna be doing something mathy. 206 - > Um I like one of my earliest memories is probably uh we went 207 - > on a family holiday and I packed a like my math textbook. 208 - > Um and this wasn't this wasn't a textbook from school. 209 - > This was one that I bought for fun.

210 - > Like that's and that was that was what I did. 211 - > Like I just really found joy in doing my times tables. 212 - > Um I that's just what I did for fun. 213 - > Um, so I think if if you could go back to me when I was six and 214 - > say, is this what you think was gonna happen?

215 - > I'd probably go, yeah, that sounds about right. 216 - > Like that tracks. 217 - > Shannon Kempe: Um I love that. 218 - > Um, I can't, yeah, I again I can read that at so many levels.

219 - > But uh so uh so okay, so you math as your is your jam. 220 - > So you start growing up, you start going through school, you 221 - > know. 222 - > Um where do you start? 223 - > Do you keep going with the math or where do you start studying?

224 - > SPEAKER_02: Um so I was like a 90s kid. 225 - > Um, I'm gonna age myself. 226 - > So I was a 90s kid, and and that was when computers, I think, 227 - > were really starting to get big from a residential, like from a 228 - > home perspective. 229 - > Um, and my parents were really supportive of what I was doing.

230 - > And and my mom, uh she's a self-confessed technophobe, like 231 - > she hates things with circuits in. 232 - > Uh, and she she said, she's like, I don't get it, I don't 233 - > like it, but this is what it has to be. 234 - > And my mom and dad saved up and got me my first computer, and 235 - > that's where I got the bug for um for for programming, is is 236 - > because you know, programming is applied logic. 237 - > They saw uh that I did a lot of, you know, when whenever we went 238 - > to my grandma's house, uh, my grandma had like a Commodore 64 239 - > that she was very proud of.

240 - > And and what I would do is we'd go up like once or twice a 241 - > month, and during the month, I'd get my textbooks and I'd write 242 - > everything down on paper in the 90s, by the way, not not that 243 - > long ago. 244 - > I'd write all of my computer programs down on a piece of 245 - > paper, and I'd I'd have to think them through. 246 - > Uh, and then I'd I'd go, you know, we'd go to grandma's 247 - > house, and as soon as we were finished with lunch, I'd run in 248 - > and I'd start timing things out and I'd see what happened.

249 - > And you really get an appreciation. 250 - > Like I think we don't realize how great computers are because 251 - > you get a really big appreciation for making sure 252 - > you're not writing that many bugs when you're not going to 253 - > get another chance to run it for like two weeks. 254 - > So that was that was, I think, the next big transition was like 255 - > getting onto computers and going, this is like math, but 10 256 - > times faster. 257 - > Shannon Kempe: Indeed.

258 - > So as you progress, though, as you get, what's your um as 259 - > you're navigating through school? 260 - > Are you studying computers? 261 - > And and as you go um it can start choosing what courses you 262 - > take. 263 - > What do you, what is, what are you leaning towards?

264 - > Where's your first job going ahead and how are you getting 265 - > into that? 266 - > SPEAKER_02: Well, uh, so I did some computer school other 267 - > classes. 268 - > Um, I really flunked out of English, which kind of limited 269 - > what I could do when I graduated. 270 - > So naturally, the very next thing I did was I went into 271 - > nursing.

272 - > Um, so I I I I was like, well, I got in and I'm gonna give this a 273 - > shot. 274 - > So I did that for a little while and I was reflecting on it uh in 275 - > preparation for this, and and there was a story that I like to 276 - > tell because I'm I'm actually not a big fan of you know 277 - > over-digitization of systems. 278 - > And one thing that I remember was uh when we're in nursing, uh 279 - > it's very data-driven. 280 - > It's it's very data-driven.

281 - > You go around once an hour and you take everybody's pulse, you 282 - > take everybody's temperature, you'd take everybody's uh blood 283 - > pressure, and you'd have to write it down. 284 - > And on the sheet, we we you know, every single person, we 285 - > would write them, you know, we'd we'd not only write it down, but 286 - > we would actually plot it hour by hour. 287 - > Um, we'd plot how fast their breathing was, how fast their 288 - > their heart rate was. 289 - > And you know, if somebody was you know coming out of recovery 290 - > or if they were like in surgery or if they were really unwell, 291 - > you might do that once every 15 minutes.

292 - > And what they told us was look for the trend line. 293 - > Like when you're when you're writing it down on a piece of 294 - > paper 15 minute by 15 minute, and you're starting to see 295 - > someone's pulse rate go down, like and you're drawing it with 296 - > a pen, and you're starting to see someone's respiration, like 297 - > their breathing rate go down, and you're drawing it down, and 298 - > and it's very visceral, and you're like, something's wrong. 299 - > Like it's very hard to skip over that.

300 - > It's very hard to miss it. 301 - > And and that's what I think we're missing now, is like that 302 - > visceral attachment to the data. 303 - > I think when you're drawing it out, when you're when you're 304 - > having to make a graph, you know, with pen and paper, as 305 - > archaic as it sounds, you start to get a feeling for what you're 306 - > looking at. 307 - > Like you start to feel, like you start to to get an intuition as 308 - > you're going through about what you should be looking for.

309 - > Shannon Kempe: Don't disagree with that. 310 - > Somebody who prints things out on the regular to observe it and 311 - > write it out and to really make sure I'm I'm grasping it and 312 - > retaining it. 313 - > So nursing. 314 - > Well, what made you go towards nursing?

315 - > SPEAKER_02: Um, my mom was a nurse. 316 - > And like I said, like uh a lot of the the major um like degrees 317 - > that I wanted to get into required a higher score of 318 - > English for the universities. 319 - > So like my backup was I was like, my mom was a nurse. 320 - > Um my my um a couple of my aunts were nurses, my grandma um you 321 - > know did some some healthcare stuff.

322 - > So I'm like, well, that's what I'm gonna do, I guess. 323 - > I'm just gonna give that a shot. 324 - > And I gave that a shot for a couple of years. 325 - > Um uh once I got a chance, I actually pivoted back into 326 - > computing.

327 - > But you know, it was it was you I learned a lot. 328 - > Like there was a lot of reading, a lot of writing, a lot of 329 - > research in the in um health informatics. 330 - > And it gave me just, like I said, like a very different 331 - > appreciation for what data means because it's you know, I think 332 - > it's it's uh we we get detached from what data's about. 333 - > And and when you're drawing and when you're recording the data, 334 - > and you know, you're looking a person in the eyes, and you can 335 - > see that like this is the they just kind of relate one to one.

336 - > Uh, you you start to understand why data matters. 337 - > Shannon Kempe: Yeah, absolutely. 338 - > So, okay, so you're you're in nursing, and you mentioned that 339 - > you're getting into computing again. 340 - > So, how do you start that pivot?

341 - > SPEAKER_02: So that was just a change of degree. 342 - > I changed my degree and I kind of moved across. 343 - > So I moved campus and I just kind of started a campus and you 344 - > know was going through that degree. 345 - > Um, and and that was that was a bit, you know, that was just 346 - > like my next next natural progression was just getting 347 - > into software design, software engineering.

348 - > Um, when I was doing that, probably my my my fun story 349 - > there is as I was doing that, I I got a a job uh through my dad. 350 - > He helped me get a job in a warehouse of all places. 351 - > And I'm probably one of the few people you've spoken to who 352 - > worked in a data warehouse. 353 - > Uh not not with a data warehouse.

354 - > I worked in a data warehouse. 355 - > Physical warehouse, yeah. 356 - > It was it was uh so so it was a a massive warehouse filled with 357 - > uh tapes, you know, of of seismic data uh from before I 358 - > was born. 359 - > And my job was taking them from the box, like going into the 360 - > warehouse, going into the data warehouse, finding the data box, 361 - > getting the data tape, putting it onto a machine that was built 362 - > before I was born, and then helping to dig digitize it 363 - > across into a new archival format so it wouldn't rot.

364 - > Like it was physically rotting. 365 - > Um, these tapes were physically rotting. 366 - > Uh, so we actually had to do a lot of weird things to get them 367 - > to work. 368 - > So we had to bake them.

369 - > So we'd take these old tapes and we'd be putting them in the oven 370 - > to kind of cook and soften the tape so we didn't strip the 371 - > magnetic tape. 372 - > Because this is data that's been recorded in the field, like it's 373 - > been recorded on a truck in a dusty, humid environment, you 374 - > know, exposed to the elements. 375 - > And we're having to go through and go, oh, that data's from, 376 - > you know, that was data collected in the tropics, so 377 - > that's probably quite wet.

378 - > We're gonna have to bake it a little bit longer. 379 - > So we were we were doing all sorts of weird, funky science 380 - > just to try and get this out. 381 - > And that was probably the time when I learned about why 382 - > metadata was important as well, is because you know, we would 383 - > have to go into the warehouse and we have to find a box of 384 - > data. 385 - > So, you know, we can't just go in and hope we're grabbing the 386 - > right, the right, you know, chunk, the right tape.

387 - > You know, everything was labeled, you know, where it was 388 - > from, where it was collected, how it was collected. 389 - > You know, there was a little information sheet that just kind 390 - > of sat on there, and uh, and that was what gave us the 391 - > information we needed to be able to grab it out effectively. 392 - > So it was, you know, probably without even realizing it was my 393 - > first taste into why metadata was useful, because I I can tell 394 - > you this like when when the tape, like when you come across 395 - > a palette of of tapes and all the labels have fallen off, when 396 - > you have to open them up and try and rematch everything, you you 397 - > wish that there was good metadata on the outside of that 398 - > box.

399 - > I'll tell you what. 400 - > Shannon Kempe: Yeah, I'm sure. 401 - > Wow, a little bit of a mystery there. 402 - > Um that's I've never heard of uh of making the tape to uh to save 403 - > it.

404 - > It was nice. 405 - > Yeah. 406 - > Yeah, that's fascinating. 407 - > All right, so you're working in uh warehouse, data warehouse.

408 - > You're finishing up uh a computer degree. 409 - > So where are you going next? 410 - > SPEAKER_02: Where am I going next? 411 - > I don't know.

412 - > Like I well, I was I was kind of working at a couple of jobs. 413 - > I was bouncing around. 414 - > I did a I did a lot of odd jobs when I was younger. 415 - > Um, probably the next big transition was towards the end 416 - > of my degree.

417 - > I was I was just working a call center job, and I found out 418 - > about like a 12-week internship at at the the Australian Bureau 419 - > of Statistics. 420 - > And I was, I'd I'd been studying for like six or seven years, 421 - > bouncing around degrees, trying to figure out what I was gonna 422 - > do. 423 - > And I was like, I like I like data. 424 - > I like data, like statistics is good, government job is good.

425 - > I I think that's what I'm gonna do next. 426 - > I'm gonna give that a shot. 427 - > So I apply for this internship, this 12-week internship. 428 - > And it's actually how I met my colleague here in Wisconsin is 429 - > through through the story.

430 - > Uh, so I was given I was given this project to work on. 431 - > Like they they come in and they're like, do this project on 432 - > this weird metadata thing. 433 - > And and I think this was the first time I heard the word 434 - > metadata. 435 - > They go, there's this weird metadata format.

436 - > Um, can you can you do some intern stuff and see what it 437 - > means? 438 - > I'm like, okay, I'll give it a shot. 439 - > So I've never worked with XML before, and I'm just trying to 440 - > hacking away. 441 - > And and and what it was was it was a standard called the Data 442 - > Documentation Initiative.

443 - > They're still around and they do a lot of work on um 444 - > questionnaire design. 445 - > So they have this big XML format for questionnaires. 446 - > And I'm like, you could just convert the metadata into the 447 - > questionnaire. 448 - > Like, you don't even need to do both, you could just do one.

449 - > I'm very lazy. 450 - > I'm like, I just could do half as much work. 451 - > What if we design the metadata and then just make the the 452 - > survey from that? 453 - > And people looked at me, they're like, You can do that?

454 - > And I'm like, Yeah, look, here, here's an example. 455 - > And they fly me to Canberra, the capital of Australia where I 456 - > live now, and and they said, We're gonna fly you in for two 457 - > weeks to meet this consultant who a gentleman by the name of 458 - > Arafon Gregory, he's watching, uh, probably was the start of my 459 - > career. 460 - > As this guy uh was in Australia and he was you know helping with 461 - > metadata for the Australian government. 462 - > And they they fly me across to kind of do a little demo, and he 463 - > looks at it and he's like, You're gonna have to present at 464 - > my conference.

465 - > And everybody's kind of stunned because the conference is like 466 - > six months away, and I'm meant to be like my my contract's up 467 - > in about four weeks' time. 468 - > So they they kind of you know give me a uh uh you know a new 469 - > part-time gig. 470 - > So I do that for the for the for my final year of university, and 471 - > halfway through the year I have to apologize to one of my 472 - > professors, and I'm like, well, I've I've got to go over to the 473 - > US.

474 - > And he's like, Well, we let people skip exams for sport all 475 - > the time. 476 - > This is much more important. 477 - > Okay, fine. 478 - > And uh, yeah, so to go, I have to get a passport because I've 479 - > never traveled anywhere.

480 - > Um, and uh yeah, so I fly across to Cornell University, having 481 - > never left Australia, having never really gone anywhere at 482 - > all, I fly over and I present this research, and everybody's 483 - > like surprised that we could do this. 484 - > And that was what kind of gave me the bug because I think it's 485 - > it's um today still, I think uh the concept of metadata is still 486 - > um poorly understood. 487 - > Uh, and I think um, you know, we don't know why we do these 488 - > things.

489 - > And that's what has always interested me is how do we help 490 - > people understand the things that we know are important, you 491 - > know. 492 - > And I know you and I have had some conversations about this 493 - > recently is, you know, we know that data is important. 494 - > We know that data governance and data management are important, 495 - > um, but we we struggle to get that attention elsewhere. 496 - > And that's what I what really fascinates me is how do we get 497 - > everybody else to kind of understand what we're doing?

498 - > And that's that's kind of brought us up to today is like. 499 - > A lot of my role now is, you know, as I said, like for me to 500 - > get product into market, I have to convince executives um why 501 - > this is important. 502 - > And, you know, to do that, we need to talk about value and 503 - > say, well, you know, if we do this, we're going to get better 504 - > data and we're going to get better outcomes and we're going 505 - > to get it faster. 506 - > Um, and I I think uh people's on that with without each of those 507 - > steps, without knowing why data mattered to individuals, why 508 - > metadata mattered to people, you know, digging through the data 509 - > warehouse and how to communicate that to people so that they 510 - > really pay attention.

511 - > SPEAKER_01: We get this question a lot in our webinars. 512 - > How do I improve my data governance skills? 513 - > Shannon Kempe: Data diversity is applied data governance 514 - > certification. 515 - > It turns what you're learning in our free sessions into a 516 - > credential that actually signals expertise.

517 - > SPEAKER_00: If you're leading or launching data governance, 518 - > search applied data governance certification from Dataversity 519 - > and see what's inside. 520 - > Shannon Kempe: Um next after you've given this presentation, 521 - > you graduate. 522 - > SPEAKER_02: Oh, so where do I go next? 523 - > So um, well, I guess the next one from there is uh so I kind 524 - > of hopped around the public service, which was super fun.

525 - > And um the probably the the next place that I landed was uh at a 526 - > the Australian Institute of Health and Welfare, and I was 527 - > like a solo programmer on their metadata system, their metadata 528 - > platform. 529 - > Like we need someone to come in. 530 - > And a colleague of mine recommended me in, and I'm 531 - > working there, and I still don't quite get it, to be honest. 532 - > Like it's it's only like in the maybe the last six months that I 533 - > actually understand why metadata is important.

534 - > Um, I've only been running Aristotle for like 10 years, and 535 - > it's only in the last maybe year and a half I've really 536 - > understood why the why everybody cares. 537 - > Uh so I'm working on a metadata registry, and we keep getting 538 - > these emails in, and we get an email like once a month, and you 539 - > know, they're coming through to me because I'm I'm the software 540 - > guy, and it's like you know, health departments around the 541 - > world saying, Can we have this platform?

542 - > Can we have this platform? 543 - > And I'm like, this is kind of weird. 544 - > Like, isn't there isn't there something else? 545 - > And the one that I remember is I think it was the Croatian 546 - > Ministry of Health, I think it was them that sent an email and 547 - > they said, We need one of these.

548 - > How do we get one? 549 - > So I turned to my boss at the time and I'm like, what do we 550 - > do? 551 - > Like, I'm really big into open source. 552 - > Can we can we give them this?

553 - > How do we give them this? 554 - > And they said, No, no, no, we're a government department, we're 555 - > not a software company. 556 - > And I'm like, okay. 557 - > And you know, as I was learning about what this system was, it 558 - > was all based on like a nice, open, freely available ISO 559 - > standard.

560 - > And I went, I can, I'm pretty sure I can build one of these. 561 - > And and that was what I I learned after the fact, it's 562 - > what they call product market fit. 563 - > When when people come to you and go, we want this thing. 564 - > Uh, so that's what I did for for maybe a few years, is I just I 565 - > just spent time hacking and building this this metadata 566 - > registry and releasing versions into the world to see who was 567 - > interested.

568 - > Um we stopped open sourcing uh a few years back for commercial 569 - > reasons, but we were just, I was just like, well, obviously 570 - > somebody wants this. 571 - > Let's see what happens. 572 - > Um, and then along the way, um, I met my co-founder who was 573 - > traveling through Australia um to who really helped you know 574 - > start that communications journey. 575 - > So Lauren, my co-founder, is is really responsible for the you 576 - > know the communication side, so that I was able to focus on the 577 - > technical side.

578 - > And and then it's just kind of snowballed from there. 579 - > Like I think a few years later we were able to land some 580 - > clients that meant we could both you know quit our jobs and do 581 - > this full-time, and and we've just kind of continued to to 582 - > grow. 583 - > I think like I think um, I don't think I'm at the end of my 584 - > career. 585 - > Like I don't I I I think I've probably interviewed for my last 586 - > job.

587 - > I don't think I'm gonna get better than the CEO. 588 - > But I think there's still so much to be done around you know 589 - > getting people to really appreciate like why why metadata 590 - > is valuable and why data governance is valuable for for 591 - > you know for companies trying to increase their value. 592 - > Shannon Kempe: So you mentioned earlier and that I want to come 593 - > back to the you know, you've only learned the true value of 594 - > it in the last year and a half, which you know, I think you've 595 - > understood it a little bit longer than that.

596 - > But but what's really driven that uh that what's driven that 597 - > statement? 598 - > SPEAKER_02: Um I think what's what's driven that is one of the 599 - > things that I did do is is I uh about four or five years ago, I 600 - > started an MBA, I finished my MBA about a year and a half ago, 601 - > I guess. 602 - > And I did that because I I always had a technical 603 - > background, I always had a lot of technical training. 604 - > And my rationale was, uh, and I looked into this for a book that 605 - > I'm writing on on the MBA journey, is I looked into it and 606 - > it's I think it's you know, 40% of CEOs, or 40% of CFOs, chief 607 - > financial officers, um, and I think 50% of CEOs uh in in the 608 - > US at least have MBAs, have Master of Business 609 - > Administrations.

610 - > So I said to myself, that's what I need to learn. 611 - > Like if I'm gonna talk to these people who hold the purse 612 - > strings, I need to know the same language. 613 - > Uh, I need to know how they talk, I need to know how they 614 - > value things, I need to know what they do. 615 - > Like I need to know, you know, how when they look at a 616 - > business, what do they see?

617 - > I think the the biggest challenge that I saw was um, you 618 - > know, in an MBA program, you learn a little bit of marketing, 619 - > you learn a little bit of finance, you learn a little bit 620 - > of accounting, you learn some strategy, um, you know, you 621 - > learn HR things, but we never covered data really to a to a 622 - > big depth. 623 - > And I've looked at a lot of MBA programs, and I think that's 624 - > that's probably the big opportunity for you and I is 625 - > there's not a lot of depth of knowledge at the executive level 626 - > to help them understand why data was important.

627 - > And and that's why when I say in the last um year and a half, 628 - > that's where my focus has really been is you know, we know, like 629 - > like I've known for a while, like metadata is important 630 - > because it's just useful, it helps us do things, it helps us 631 - > find things, it helps us understand things, but it's 632 - > really being able to quantify and talk to that business 633 - > imperative of going, this is good for business, and and being 634 - > able to say why.

635 - > And and uh, you know, we're we're conducting some research 636 - > at the moment, it's still preliminary, and I don't want 637 - > to, I don't want to, you know, talk too much without letting 638 - > our partners kind of release this out there. 639 - > But what we're seeing is, you know, data governance and 640 - > metadata are a multiplier for business value. 641 - > Like they're not nice to have. 642 - > Like we've actually, we we recently did a survey with 130 643 - > participants, and we found that there's a direct correlation.

644 - > There is a really strong link between, you know, data value, 645 - > like between you know, good data governance and and good 646 - > documentation of data and how people perceive value in their 647 - > organizations. 648 - > And and that I think is the thing that we need to hit on is 649 - > to say, we're not doing this because it's nice. 650 - > We're not doing because it's really fun philosophically. 651 - > It's really fun to define customer.

652 - > It's really fun. 653 - > Like you probably did you have a lot of fun, like, you know, when 654 - > you define a customer. 655 - > Like, was it was it worthwhile? 656 - > Like, did you really enjoy it, Shannon?

657 - > Shannon Kempe: Uh well, I do. 658 - > I had fun working. 659 - > We just so for everybody, we had a team exercise where we defined 660 - > customer for data versity. 661 - > We were talking about it pre-show.

662 - > Uh, and yeah, so I mean, it was fun for me to experience the 663 - > conversation and to have the, you know, to ensure to 664 - > experience, you know, what everybody goes through, right? 665 - > SPEAKER_02: And I think like like I think a lot of people 666 - > listening will just go, it's it's just fun to define 667 - > metadata. 668 - > It's kind of fun to argue about how we're gonna categorize or 669 - > classify these things. 670 - > Um, but I think that's where what I what I want to do is take 671 - > a step back and go, well, no, no, no, it's not just fun, it's 672 - > not just interesting, it's not just intellectually challenging, 673 - > it's delivering business value.

674 - > Like it's it's we're doing this because it makes organizations 675 - > more effective and more efficient. 676 - > And I actually like as part of this research, have been looking 677 - > into the data that we collect about our own industry, and it's 678 - > the it's actually surprisingly little. 679 - > There's actually surprisingly little evidence around why we do 680 - > the things we do. 681 - > And I think it's just because we just believe it to be true.

682 - > Like nobody needs to explain why data governance is good, we just 683 - > know it. 684 - > And and you know, from my perspective, when I'm trying to 685 - > convince somebody to to invest a lot of money in a program, I 686 - > have to talk, you know, about that business value. 687 - > And and that's what I think the next the next big leap for for 688 - > for data governance at data management is speaking to that 689 - > imperative, saying it's uh it's just good for value, it's good 690 - > for value, it's it's good for generating value, it's good for 691 - > generating trust.

692 - > Shannon Kempe: Um, how do you think the uh AI boom is uh 693 - > affecting that? 694 - > SPEAKER_02: I think uh, well, I I think you know data and AI uh 695 - > uh go hand in glove. 696 - > Um you can't do AI without good training data. 697 - > You need data to feed into it.

698 - > And we've seen a bit of interest and a bit of uptake in people 699 - > realizing that you know you can't do you can't automate data 700 - > and you don't have metadata that describes what it is. 701 - > I've seen people go, oh, it's context. 702 - > It's like it's metadata, we've had it forever. 703 - > Um, so I think that's that's driving it somewhat because AI 704 - > is helping make really bad decisions really fast.

705 - > Like it's it's really great. 706 - > I saw one today that you you can't Google for the word uh for 707 - > the word cancel anymore. 708 - > If you Google the word cancel, the AI prompt just says, okay, 709 - > I've stopped working. 710 - > And if you Google like disregard, Google just says, 711 - > okay, I won't do anything for you.

712 - > Uh so it's that kind of thing where where AI is AI is really 713 - > powerful when it has the right data and when it has the right 714 - > context. 715 - > And that I think is that it that's where that value comes 716 - > from, is you know, our ability to get those insights quicker, 717 - > but it still requires good governance at the head. 718 - > Absolutely. 719 - > Shannon Kempe: Uh uh, so tell me, Sam, what has been your 720 - > biggest lesson so far in your career?

721 - > SPEAKER_02: Um uh I think I was gonna go with something boring, 722 - > like always ask why. 723 - > I I think I'm gonna go with something a little bit more 724 - > controversial and say know when to push buttons, right? 725 - > No, know when to push buttons. 726 - > Um, you know, I've made a a career of making a lot of really 727 - > good mistakes.

728 - > Um I went, you know, well, I was talking about that story where I 729 - > was invited to a conference. 730 - > I was invited to a conference because a room full of senior 731 - > executives it said, you know, this is interesting, but we just 732 - > can't do it. 733 - > And I literally grabbed the monitor cord and I jammed it 734 - > into my laptop and then I said I did it over the weekend. 735 - > Like we can do this.

736 - > Um, you know, that probably, you know, won me from some friends. 737 - > It probably lost me some friends, but it it got 738 - > attention. 739 - > Um I've and I've made, you know, I've made some good mistakes, no 740 - > bad, some bad bad mistakes were, you know, there's good 741 - > attention, there's bad attention. 742 - > The one at the moment is um, and I think I I this is what I 743 - > presented with Dataversity, is I was presenting on a on a data 744 - > program here in Australia.

745 - > The Australian government has been you know investing a lot of 746 - > money, like$200 million on a on a data inventory platform to 747 - > improve discovery of data. 748 - > And you know, over that$200 million, over four years, uh, 749 - > we've seen 34 data requests. 750 - > We've seen 500 assets found, and and the quality was really poor. 751 - > And I made a decision earlier this year is I'm gonna talk 752 - > about this.

753 - > I'm gonna show people what's happening. 754 - > And that meant, you know, writing down who was doing 755 - > really poorly. 756 - > This was, you know, talking about the the Australian tax 757 - > office having one data, you know, and and this is what I 758 - > presented at Dataversity is I said, like, do you think it's 759 - > right that the tax office has one data? 760 - > You know, there's and then a lot of people say, like, like, well, 761 - > like Shannon, do you think it's okay that the tax office has one 762 - > data?

763 - > No. 764 - > Like we argue it's not possible. 765 - > It's not possible, it's not right. 766 - > Like, that's you know intrinsically.

767 - > I don't need to go to I we don't need to define data. 768 - > We don't need to define data to know that the the tax office has 769 - > more than one data. 770 - > There's more than that, right? 771 - > And and I've I've repeated that story a lot.

772 - > They still only have one data. 773 - > Australian tax office, if you are listening, you still only 774 - > have one data, and I'm gonna keep pointing that out. 775 - > Um, and that's that's very confrontational. 776 - > Like it's very confrontational to to you know to take that out 777 - > and and write that down.

778 - > And it was it was harder because I was invited to present that on 779 - > on live TV. 780 - > And that that only elevated how much uh people saw it. 781 - > Uh, but but by taking that risk, um, we've actually seen 782 - > improvement. 783 - > We've actually seen a 1% improvement where we haven't 784 - > seen improvement for nearly 12 months.

785 - > Um, people are actually going back. 786 - > Like, you know, chief executives of of uh you know government 787 - > departments are actually encouraging people to do better 788 - > because now we've shone a light on them and said, well, there's 789 - > a reputational risk at play if you don't actually do data 790 - > governance properly. 791 - > And and I think that's probably the hardest lesson to learn is 792 - > to know when to push the button, like to know when to to take a 793 - > risk.

794 - > And and uh I I've heard someone say that you know company 795 - > founders are actually exceptionally risk averse 796 - > because you know what's at stake. 797 - > Um, they're very risk-aware. 798 - > And I think that's very true. 799 - > And I think the the challenge for anybody at any point in 800 - > their career is figuring out what is the risk you can take 801 - > and what's the what's the payoff and what's the you know, what's 802 - > the the the risk.

803 - > Um and I think you know, figuring out what your risk 804 - > profile is, figuring out the chances you want to take helps 805 - > you understand how you can make those big evolutionary leaps in 806 - > your career. 807 - > Shannon Kempe: I absolutely agree. 808 - > Um and uh you're not always gonna succeed, of course, right? 809 - > But no, you'll learn from the you'll learn from the mistakes.

810 - > SPEAKER_02: Uh let me see. 811 - > I what was the biggest risk that I took that it didn't pay off? 812 - > I think um we worked with a client uh quite a long time ago 813 - > who who um you know we worked quite a long time on this 814 - > contract with them, and at the end it the the relationship kind 815 - > of fell away, and they actually tried to take our open source 816 - > code, like they actually tried to to take you know all of our 817 - > code and all of our commercial code would have been quite 818 - > ruinous for the company, and we had to kind of fight to to keep 819 - > that and fight to retain it.

820 - > And and that was probably one of the biggest risks we took 821 - > because it was very, very early on was working with a very large 822 - > client. 823 - > Um, so I think it's you know, you you you take the good with 824 - > the bad, you know, you it's uh I'm always the worst at 825 - > listening to my own advice because you forget these things 826 - > along the way. 827 - > But you know, when when you take a risk and when it doesn't pay 828 - > off, I think you just got to listen to it and go, what did 829 - > what did I learn here?

830 - > And then how do I make sure that I can not have that happen 831 - > again? 832 - > Shannon Kempe: Yeah, absolutely. 833 - > Okay, so Sam, so tell me, do you see the importance of data 834 - > management and the number of jobs working with data 835 - > increasing or decreasing over the next 10 years and why? 836 - > SPEAKER_02: Um I think uh I think we're probably gonna see 837 - > an increase in jobs.

838 - > Um, I think uh we're gonna see more people involved in data 839 - > governance. 840 - > I mean, I'm I'm making sure of that, right? 841 - > Like when we when we when we drive attention onto this, when 842 - > we say, you know, data governance improves value, we 843 - > want to see investment in that, not just in tooling, but in the 844 - > people to run those tools. 845 - > You know, I think um, and I think what we're going to see as 846 - > well is I I hope we see a very similar effect to what we saw 847 - > with Jira.

848 - > Like, I think Jira is an as an Australian company, you know, 849 - > Jira and Atlassian are a natural um company for us to want to 850 - > emulate. 851 - > You know, they were an Australian firm, they were 852 - > focused on business software, and they've been very, very 853 - > successful because of it. 854 - > Um, Mike and Scott, really, really eager to chat, but um, 855 - > you know, we'll see how that goes. 856 - > Um but I think the other thing that we saw from that was a 857 - > democratization of project management.

858 - > You know, project management went from a really, really niche 859 - > skill to all of a sudden everybody was an agile project 860 - > manager. 861 - > And and I think that was a good thing because uh, you know, you 862 - > don't need to be an expert in project management to run a Jira 863 - > project well. 864 - > Like you do your agile you know, sprints, you do a retro, you do 865 - > some Kanban boards, and you're probably gonna be pretty 866 - > effective. 867 - > And and I think we need to figure out the same for data 868 - > governance is there's always gonna be a role for senior 869 - > experts, you know, really experienced veterans.

870 - > Um, but I think when we can when we can really you know put into 871 - > practice, I think Rob Seiner said it best, um, the the author 872 - > of non-invasive data governance, he says everyone's a data 873 - > steward. 874 - > Uh I think you know, if everyone's a data steward, we're 875 - > gonna see massive growth in data jobs. 876 - > But I also think there's there's gonna be a big role for you know 877 - > that equivalent of our PARA project managers, you know, 878 - > people who are you know a little bit, you know, a little bit 879 - > beyond where they're at now, understanding how data 880 - > governance is good from a from a from a you know from a business 881 - > perspective.

882 - > So I think I think I think the the the growth of early career 883 - > opportunities and the growth of you know democratization of data 884 - > management and data governance is going to be really important. 885 - > And I think the reason that that that has to happen is because 886 - > everybody's making data. 887 - > You know, you can run a you can run an Eventbrite survey and or 888 - > you can run an Eventbrite event, you put you know, you put it up 889 - > there, you start asking questions, you're collecting 890 - > data.

891 - > You go to SurveyMonkey, you can run a survey, you can put it 892 - > online, and you can get a thousand, two thousand 893 - > responses. 894 - > You know, you put a couple of ads behind it, you get thousands 895 - > of responses. 896 - > Um, we're all making data. 897 - > And I think we need to understand how we can get 898 - > everybody involved in data management and data governance 899 - > so that we continue to get value out of those assets.

900 - > Shannon Kempe: I agree. 901 - > So, what advice then would you uh give to people looking to get 902 - > into a career in data management? 903 - > SPEAKER_02: I think um I I would really, really stress the 904 - > importance of soft skills in quotes. 905 - > Uh, I think you know what we saw in that research was when people 906 - > talk about why their data is important, they got more value.

907 - > And why can't come from a machine learning system? 908 - > You know, it's not going to come from AI. 909 - > Why did you collect this data only comes from the business 910 - > people who made it? 911 - > Why did you collect dietary requirement?

912 - > Because I didn't want anybody to die when they came to the 913 - > conference. 914 - > You know, I wanted to get a list of allergies, because we we 915 - > forget about that, right? 916 - > When when you collect, you know, there's two questions that I 917 - > always see at a conference. 918 - > You ask, you know, do you have a dietary requirement?

919 - > And do you have any allergies? 920 - > Now, you may not be aware of it, but when you ask about diet, 921 - > what you're really asking is you're asking about religion. 922 - > Right? 923 - > When you ask, like, you know, vegan, vegetarian, okay.

924 - > But when you ask, you know, are you kosher? 925 - > You know, are you hawal? 926 - > Are you Jain? 927 - > You know, like you've just you've got a proxy measure for 928 - > someone's religion.

929 - > That's a very sensitive piece of information. 930 - > And the same for analogy, that's medical information. 931 - > Like, we we need to know why we're doing that. 932 - > Why did you collect it?

933 - > Because I didn't, I wanted to serve them the food that they 934 - > were allowed to eat and I didn't want to kill anybody. 935 - > You know, all of a sudden we have a rationale for for having 936 - > two very sensitive pieces of information. 937 - > So I think um, you know, I would I would encourage those soft 938 - > skills and and I think that the the opportunity for us as a as a 939 - > as a as an industry to lean into philosophy majors and psychology 940 - > majors who are both really used to asking those questions why 941 - > and doing doing data research, I think is it's it's an untapped 942 - > gold mine, to be honest.

943 - > Like it's an untapped gold mine of resources of people who are, 944 - > I think, you know, already trained and already well 945 - > equipped to be the data governance and data management 946 - > leaders of tomorrow because they're the people who will help 947 - > us make sense of why we're collecting all of this data to 948 - > begin with. 949 - > Shannon Kempe: I really believe in that philosophy, which is why 950 - > we've you know made sure that we offer communication training 951 - > along with our data management training.

952 - > Yeah. 953 - > It just is so important. 954 - > And a skill that we can always work on, right? 955 - > That's get better at.

956 - > Yeah. 957 - > SPEAKER_02: Somebody asked me if I got media training recently. 958 - > The answer is yes, because you have to get better. 959 - > You know, I think I and I think um we we can't rely on other 960 - > people wanting to talk to us and and learning how we talk.

961 - > You know, the data profession is is there's jargon because it 962 - > helps us talk to each other really quickly. 963 - > But I think we need to understand how we can break down 964 - > those walls and how we can bring bring the impact of what we do 965 - > because because they're not gonna listen to us. 966 - > Nobody's gonna invite you into the boardroom. 967 - > You have to kick your way in.

968 - > Like I see a lot of people online complaining that nobody's 969 - > bringing data into the boardroom. 970 - > It's like, well, stop asking for permission. 971 - > You're you're there for a reason. 972 - > Go in.

973 - > Shannon Kempe: I just want to expand just a little bit. 974 - > One of the reasons I was really excited to hear about your 975 - > career because I knew we talked a little bit about the 976 - > conference, about how you started in as as in nursing uh 977 - > in your schooling. 978 - > We actually have, and what's something I didn't mention is we 979 - > actually have a lot of um nurses who have are going into data. 980 - > They are transitioning because they can no longer handle the 981 - > physical demands of nursing, but are not ready to retire.

982 - > So they're switching into careers and data. 983 - > So like it's which is fascinating and I love it so 984 - > much. 985 - > And so do you have any advice for for that community making 986 - > that transition? 987 - > SPEAKER_02: I I won't name them because they're very private 988 - > people, but uh it's actually quite a coincidence because um 989 - > one of my probably my my chief expert on metadata uh is was a 990 - > nurse, was a was a surgical nurse, and somebody who helped 991 - > with the design of our recent maturity survey also had a 992 - > background in health informatics and nursing.

993 - > Uh so it was it was quite an odd coincidence that we all got 994 - > together and we all started talking about you know how we 995 - > got to where we are. 996 - > And I I think it's for that reason, is like it's you 997 - > interact with data on the daily. 998 - > I think uh for anybody who's considering pivoting from 999 - > nursing into data, I think do it. 1000 - > I think uh you're probably more familiar with the impact of data 1001 - > than other people.

1002 - > Like when you see when you see someone has a as a has a pulse 1003 - > rate of 20, you don't need to explain what that is, you just 1004 - > go, oh, that's bad. 1005 - > And when you start to see those trends, you you know that's bad. 1006 - > So you know why data is important. 1007 - > Um, so I would say, like, you know, if you're if you're 1008 - > looking at a career transition from nursing to data, I would 1009 - > not think that you're starting again.

1010 - > I would think that you are bringing a wealth of expertise 1011 - > in real-world data application, and you are bringing expertise 1012 - > in how data has an impact on human lives that I think would 1013 - > be invaluable to fashion. 1014 - > Shannon Kempe: I love it. 1015 - > I I totally agree. 1016 - > And thank you for that, for sharing that.

1017 - > Um, well, Sam, I would be remiss if I uh didn't ask if people 1018 - > wanted to learn more about Aristotle metadata, where would 1019 - > they go? 1020 - > SPEAKER_02: Uh Aristotlemetadata.com, of 1021 - > course. 1022 - > So we're on LinkedIn, we're on Facebook, I think we're on we're 1023 - > on TikTok now.

1024 - > You can follow our our uh our little bag. 1025 - > We've got a we've got a TikTok page set up or a I think it's 1026 - > TikTok or Instagram page set up for our luggage, uh Little 1027 - > Lucas. 1028 - > Um so that's that's probably where I'd go. 1029 - > And I'd say, um, and and also Aristotle.

cloud. 1030 - > If you're in the industry and you're looking at bringing in a 1031 - > tool to help with data governance, we actually have a 1032 - > free tool um that you can get started with all the power of 1033 - > Aristotle. 1034 - > Uh your first 10 users are free, just like Jira. 1035 - > I think it's a great model.

1036 - > So if you're interested in evaluating the system or even 1037 - > just using it, you know, because you can stay under the user 1038 - > limit, you can use it for as free for as long as you want, 1039 - > like with uh minimal restrictions to get your data 1040 - > governance really tiptop. 1041 - > Shannon Kempe: Oh, very cool. 1042 - > SPEAKER_02: Well, Sam, it has been a pleasure. 1043 - > Thank you very much, Shana.

1044 - > It's been really interesting. 1045 - > It's um it was actually uh I really like the idea of actually 1046 - > having people stat through their careers because you don't, you 1047 - > know, you just don't realize how far you've come. 1048 - > So I think it's a really great model. 1049 - > And and thank you for letting me share my very uh wonky journey 1050 - > into data governance with you.

1051 - > Shannon Kempe: Well, thank you for sharing. 1052 - > I really appreciate you taking the time today. 1053 - > So thank you very much. 1054 - > SPEAKER_02: You're welcome.

1055 - > Shannon Kempe: And thanks to all of our listeners out there. 1056 - > If you'd like to keep up to date in the latest in data management 1057 - > education, you can go to dataversity.net forward slash 1058 - > subscribe. 1059 - > Until next time, stay curious, everyone.