Aaron Phethean (00:15)
Welcome to today's show. We're joined by Nik Walker from Coop. Nik is in a typically modest leader in a big company in a big role. And we

hearing all about his technology interests, his neurodiversity interests, and him as a leader and what the future looks like as technology for data leaders. So without further ado, let's dive in.

Hello, Nik. Welcome to the show. Absolute pleasure to have you on. I'd love to dive into everything about how we met and your interests and of course, where you work and how you got into data and technology. There's tons and tons to cover. So why don't you kick us off and tell us a little bit about your role and what you do.

Nik (01:01)
Yeah, cheers Aaron, nice to be here, man. So I'm a head of data engineering at Co-op at the moment. I have a team of engineers who do kick-ass data work, shock horror, build cool pipelines, building new platforms, delivering cool insights and supporting data science and that kind of great stuff. I describe myself as a human centric leader. So this is a focus I have on treating people as individuals. It's a counseling process of...

very focused on individuals, not just their development, but in their support and where they want to go in their career, what they want to do, how to do it and basically making sure people are doing the right things. My background, I didn't even go to university. I fell into data in the most biblical of ways. It seems to be quite a common thing for many of the leaders in data at the moment to not have been to university.

Aaron Phethean (01:40)
May service.

Nik (01:46)
Maybe it's because it was a bit more meritocratic 10, 15 years ago when we fell into it. But I fell into data.

Aaron Phethean (01:50)
Do think that's changed

actually? you think there's now like, anyone coming through university thinks it's a pretty rapid improvement?

Nik (01:55)
Yeah,

I've got some quite strong opinions on the notion of productionization of data by people with MBAs who've turned what was once a bit of a startup culture could actually get loads of stuff done into basically the slowest, most laborious process known to man because it's got to focus on value. But yeah, it's always one of those things. ⁓ It's a balance.

Aaron Phethean (02:12)
you

Interesting. Well, I definitely think we should

get into that. That is the whole point. And you are typically modest and you're like, you know, you're in a big role. You're not just like, you know, a little like, you know, company co-ops massive. Why don't you tell us a little bit about co-op? mean, it's give everyone watching might not be in the UK. You might not actually know who co-op is.

Nik (02:32)
Yeah, for sure. Yeah,

for sure. So Co-op is, as it probably sounds, it's cooperative. We are a company of 55,000 people covering everything from retail. So anyone in the UK probably knows who they are because they've seen the stores. And there's the retail side, there's logistics, insurance, biggest funeral care provider in the UK. There's some people who remember the old bank. That's not part of us anymore, but it's a brand that people remember.

The schools up north, there's like 14 academies or something like that. It's just the more, yeah, the more and more you find, like the more and more I find people about, hey, can we do this thing? And I was like, what's that for? It's all for schools. Sorry, schools? What? Like I've been here 18 months and I keep finding more things. When it's an organization that's 1800s, started as a full-blown cooperative movement. It's been wholesale communities. It's all sorts of things that's gone through massive change.

Aaron Phethean (03:01)
I you mentioned that when we were chatting, but it still surprises me.

Yeah.

Nik (03:24)
this gigantic organization.

Aaron Phethean (03:25)
And

I think that was one of the things that fascinated me most when we were chatting.

because of that legacy, and I'm talking about, it's made such a big impact on the country, it's been around such a long time, there's clearly going to be a lot of history in the company and a history of technology and a history of innovation. And in a weird way, there might actually be the forefront of some things, but then really, you're really far behind in others in corners of the business. wonder where you see it, like if you look across the sort of estate that you have.

Nik (03:36)
Yeah.

There is, there's brilliant things. I talked to a sister team that sits next to mine, Data Science, and they're doing some absolutely crazy things. LLMs, AI, all the stuff you go on LinkedIn and you see day to day, you hear on Google or reading the news, it's just absolutely bonkers. You hear some parts of the business who are maybe like in need of some serious investment where they've got legacy systems that...

quite frankly, wouldn't look too bad out of a 1980s horror movie. there's a bit like a lot of people now are getting to a point where they're seeing the light and kind of modernizing that systems, that's people, people are becoming understanding of what they're doing. We can have conversations now, it's like, well, are you data driven? And they kind of go, yeah, well, we do get information from this place. And I'm like, good, you know what you're talking about. There is, there is, but there's a lot of legacy out there as well as a lot of legacy out there that's.

Aaron Phethean (04:40)
Yeah.

Nik (04:44)
slowly kind of been identified and there's this huge things in the team, the wider technical team, technology team, that data sits in has amazing, has an amazing attitude towards doing what is right for the organization and the members who ultimately own it because it is a cooperative. There's no shareholders, it's just members. We did our announcements for how much we did and how big our members are like last week and it was...

quits in, like it's really good. It's one of those things we've got like millions of members kind of thing, but it's something that's important to people. So we have a culture of like, it's good for what's good for the organization, it's good for everyone who's involved in it. It doesn't happen to be people who just work in it, it's the people who ultimately we're working for to be better for the world.

Aaron Phethean (05:12)
Yeah, really, really, really succeeding.

Yeah, that leads me to the other part that I thought you were far, far too modest about. You're in a leadership role and you're doing, I think, phenomenal things as a leader. And it strikes me that the... I mean, I've got loads of stuff in front of me that it's like just gold, all of it is like... And I think it starts with you care about people. You massively like over...

Nik (05:42)
Thanks.

Aaron Phethean (05:55)
extend on that front and that makes a huge difference as a leader. And with that many people in an organization, you know, it's I've been in big companies and yeah, become these as clients, it can be really, really hard to get anything done. So the whole people part of leading it is crucial. So I wonder, what's it like for you being a leader in co-op?

Nik (06:16)
Challenging, challenging in the right ways. Challenging in the right ways. It's a big organization. It gets some of the similar tropes that you get in there. You occasionally get politics. There's occasionally things that you have to do. There's the side of it where you have to justify things. You have to say why are you doing certain things? Why are you building certain things? Why things look the way they do to people who probably don't really understand enough what's going on?

Like it's everything that an Oracle data team does at the same time. The benefit of it is that there is a culture of absolute. We're doing the best possible value. One of the attitudes is do what matters most. And one of the core values of the organization is do what matters most. And we do what matters most because the reality is there are some people who are needing data work or tech work or

the systems that are put in place or the pipelines that are feeding into that are providing things like co-op life or funeral care systems where it speaks volumes about the kind of people who are going to need those systems at some point. Like it's, it's a big thing. So actually it's the leadership can be, it can be tough. Like you said, it's a big organization caring for absolutely everyone is, is important. I find it, it's an attitude I have to take towards building a framework and cultivating a framework towards how I behave with people. It's less about.

being the care bear and putting your arm around every person because you can. But if every conversation is safe, if every conversation is detailed and we know what we're going into and if every conversation is documented and we know what we're coming away from, it helps to make sure people, the care bear side of me can look after people without having to put my arm around someone. It forces a kind of good attitude towards it. And actually it's not an unknown attitude at co-op. We know we care about our people. That's why we work here.

Aaron Phethean (07:50)
Yeah.

Yeah, yeah. And I think it's something that most leaders would say that, you know...

It strikes me as a saying to be cruel, to be kind. You don't have to be cruel, but you might have to be realistic. You might have to be direct. You might have to be things that are not the kind of, you're not gonna dance around an issue. You have to do the right thing and do that in a caring way. And I see that coming across in what you do as well. people listening, it might not be, they don't know as well as I do. You sort of read your background and you call yourself,

Nik (08:08)
Yeah.

Aaron Phethean (08:30)
I'm reading it here. So Titan in Tweed is one of the things. Talker of mental health and neurodiversity and loser of pens and endorser of all things witty. These say a lot about you as a person and how you go about things. I'd like to dive into the neurodiversity topic because that's actually where I saw you first. I you speaking on it.

Nik (08:42)
Yeah.

Yeah, sure.

Aaron Phethean (08:52)
absolutely awesome, know, blunt and, you know, very open to say whatever you wanted as a speaker in front of a big audience. was entirely refreshing. yeah, tell us a little bit about that and how your journey and how you think you can help other people through your role or wider.

Nik (09:10)
Yeah, for sure, man. Yeah, I love talking neurodivergence more than anything because it's always one that data and tech folks relate to, but it's something that every time I talk to someone else, like, yeah, that's weirdly relatable. So I'm old guard I'm ADHD, I've got ADHD, probably some substantive autism in there, like where occasionally I'm quite socially blind to things. I describe it as social color blindness. People who get the weird into management speak at the big meetings and I'm like, I don't.

Aaron Phethean (09:20)
Yeah.

Nik (09:36)
really get what you're just be blunt about it just straight to the point it's nice and easy I don't have to dance around fuzziness so I got diagnosed with ADHD when I was like yay high when I was seven or eight many many years ago so I'm kind of old guard when it comes down to this I've always kind of advocated for myself it's been very tough over the years because it's always been ADHD neurodivergence has always been seen as like it's the naughty boys condition it's just it's what they'll be like and

I think you'll remember from the session I did, you saw some of the comments that people like myself and I've had in my life where it's like, I'll get used to asking people, do you want fries with that, like for the rest of your career? And it's like, this is stuff that people do. It's awful, but it's stuff that people do here and people don't get it. People are threatened. All you have to do is pardon the political kind of statement. All you have to do is look at the anti-vax movement where it's like, oh, vaccines cause autism. They don't.

Aaron Phethean (10:08)
Yeah.

Yeah.

Nik (10:25)
Like it's black and white. They don't, don't be an idiot about it. But like there's a genuine fear of neurodivergence, like in some way, shape or form that we are broken or we are problems. It gives me great pleasure and love to see so many more people speaking up about their needs, their wants, their experiences as a neurodivergent people person. That could be dyslexia, dyscalculia, OCD, Tourette's, ADHD, ASD, whatever we call it these days.

Aaron Phethean (10:25)
Hmm.

Nik (10:48)
how many more different things we've got under that and these people who speak up proudly as much as I do about it. It brings me such love and wonder that we're doing it finally. It's a big change because I think the joke I made was like engineers and analysts have a bit of a stereotype when there's so many of us around, like we have a bit of a stereotype that kind of comes with it, but actually it's really important to do. So it's led me to be the human being I am and lets me be the mad man I am sometimes as well.

Aaron Phethean (10:55)
Yeah, it's definitely a big change.

Well, it feels like you're in good company. The stat that I remember from the event was, I think one in 10 in the general population, you're a virgin, and three in five in ⁓ software data.

Nik (11:26)
three and five, yeah, 60 % of us, 60 % of us neurodivergent,

which can be really interesting, because obviously it causes certain dynamics, but it gets quite interesting then when you get folks who aren't, who don't really recognize that they're the odd ones out. And it sounds really bad, but it's just like, actually, you're the outsiders in our weird little world right now. it can be the difference, the cultural differences is quite interesting to see.

Aaron Phethean (11:41)
you

Yeah.

Nik (11:51)
when that happens. And I've definitely had it least once in my career where the team is mostly neurodivergent and other people are not. And you're kind of like, ooh, you are.

Aaron Phethean (11:58)
It was classic at the event, a

big room full of people. There's 150 people or so. And I definitely saw everyone going, doing a little quick, that's quite lot of us.

Nik (12:04)
What's a lot?

Yeah, it counts. Yeah, yeah, yeah,

you see it. You see it all the time. People sit there go, holy crap, that's a lot. Hang on. Like, it's like the stereotype you get. It's where, please look around the room. Nine out of 10 people will pass this class. It's very much the same thing. And you've got to sit there. People relate to it. People relate to it. And they kind of go, that's me.

Aaron Phethean (12:23)
Yeah.

Yeah, I definitely I saw people afterwards, you come up to talk to you and you're just just to engage on that level. you know, and maybe start their own journey and you start their own like, you know, their own company acceptance of their their skill set. And one of the things that I see you post a lot about

And with some, you know, think reluctance for the title, that kind of superpower, neurodiversion as a superpower. And, you know, I know you're not a fan of it. And I wonder if you could dig into that for a second. And the benefits, you know, clearly ⁓ you use your own skillset that is a benefit to whatever you bring to the table.

Nik (12:57)
Yeah, for sure.

Yeah,

I have an issue with the the with toxic positivity in general, and I often find people who describe neurodivergence as superpower to be quite irritatingly toxically positive, like, hey, everything's always good all the time. It's not. It's hard work being ADHD. Like, I take too much on I pick things up that are hard work. I do a lot of other people. It can be quite emotionally draining to be ADHD, be neurodivergent, like in one way, shape or form, or if you have to mask an awful lot, which I do sometimes.

it can be absolutely knackering, for lack of a better phrase. Like these are the disadvantages. The advantages of it are that you can focus, you think bigger than just this point. Like you'll fidget, you'll do things, but you'll figure things out. Like you've got a natural, almost a natural proclivity. don't know of many narrowed avenging people who aren't naturally predisposed to problem solving because they've had to figure themselves out. So it's natural for us.

Aaron Phethean (13:31)
Yeah.

Mm-hmm.

and

quite comfortable in getting lost in it.

Nik (13:54)
yeah, spectacularly. The amount of hyperfocus

is like 3 AM things where my wife walked in through the door and kind of going, you come into bed and I'm like, look at the time and like I've done something or like I've been reading or like I've been gaming or I've been like coding something. Like, can I just get lost in that this is just absolute hyperfocus. There's nothing else in my brain bar this thing that I have to solve or do or enjoy or whatever it is. Like it can be like that. But again, that's.

Aaron Phethean (14:02)
Already.

Yeah.

Nik (14:19)
can get overwhelming because you can get hyper-focused on multiple things. it's almost an obsession in the back of your brain that goes, you need to do this thing right now. And that can be hard to manage. That can be hard both to manage as an individual, but as a leader, that can be hard to manage when it's other people. And so there's a big education piece there that I try and try my hardest to do, to try and have that point to it.

Aaron Phethean (14:23)
Ignoring life.

Yeah.

Yeah.

And yeah, it probably helps enormously to be able to spot it and be aware of it and see that in your team. I wonder then, if we move a little bit onto tech and things that you're into, what have you been digging into lately? What's got your attention in the tech world?

Nik (14:58)
In the tech world, I am at the minute weirdly fascinated by natural language processing. Not LLMs or anything necessarily hardcore, because I have some grumpiness about AI sometimes. I am a Luddite when it comes to AI. I think AI is cool. I sit there and I see the social ramifications, and I'm like, God, I hate this. But it's one of those.

Aaron Phethean (15:09)
Probably with your tools.

Yeah.

Nik (15:21)
Natural language processing, of like from a purely technical data perspective, there's been a, it's probably less the tech itself, and more about the application thereof. One of the big things that I've been thinking of for the longest time is how irritating it is to have to turn around to seven people and say, Hey, what's going on with X data? And actually there's going to come a point very, very soon over the next few years where natural language processing takes over in a big way.

the interactivity of end users to platforms and processes like, hey, this data seems wrong. Is everything running okay that I've typed in or I vocalized and it's gone? No, one of the pipelines has fallen over. This is because SAP has failed to load something. And it tells you instantly what the problem is. And rather than just because SAP has fallen over, but it then says, I have informed Jim in SAP that there has been a problem.

Aaron Phethean (15:51)
Yeah.

Yeah.

Nik (16:11)
And this was informed with him and he'll tell you what's going on. And you don't have to have that. Like that's got my interest at the minute because that's been the missing link of tech teams for probably 30 years and data teams. we've, everything's been resolved around dashboards and reports. And actually we've been sitting there screaming about translators into data and the tech world for the longest time. It's here. It's now. And we've got the power, compute and the maturity to do it.

Aaron Phethean (16:13)
Yeah, yeah,

I feel

very similar feeling about the kind of data input and system record keeping side of things. The kind of mouse and the kind of capture of information.

It's very backwards. mean, it's really slow. really, you know, it's okay typing it in, you know, whereas I think if you, you know, observing the world listening, you know, gathering signals and stats, it's, it's the potential is amazing. So, you know, I think the actual, the game changer probably will be the natural language processing or the, you know, people think of the LMS context, understanding what's going on. That's, that's part of

Nik (17:00)
I can't believe it.

Yeah, and it's

that training towards it. Like I talked to the data science, talked to some of the data science guys at Co-op or like even friends who are data scientists. And some of the things they come out with about what they're using it for is absolutely stunning. It's absolutely just phenomenal that they've got these things. It's all trained on internal data, so only gives them back. And I think the next big thing will be the same thing that kind of data went through 10 years ago, which was people bring multiple reports to a business and like, why have we all got different numbers?

Aaron Phethean (17:27)
Mm-hmm. ⁓

Yeah.

Nik (17:36)
How do you build that confidence? How do you build that tech understanding of how do you build that trust? Like that's gonna be the next big thing for LLMs because in a minute, like I've been to a lot of conferences, I've been to a lot of places, lot of talks where people are like, these are the best things in the world. Look at this tokens, ha ha, it's amazing. Great, cool. It's still giving me bullshit back. Like it's still, I still can't, like yeah, I still can't sit there, type in and go, give me my daily sales. And it comes up and says 1.7 million. And I go, I don't think I sold 1.7 million yesterday. ⁓ yeah, exactly.

Aaron Phethean (17:53)
Yeah, because I had very poor data, really.

Yeah, not in my lemonade stand. No, that was not happening.

Nik (18:05)
And

that's, and like, yeah, okay, that's hyperbolic and like a bit ridiculous, but actually, actually that's what people worry about. if, if I, a data person worried about that, what are the C-suite going to worry about in an organization of any size? They're going to worry that it's giving them BS back and they don't want that. They want, they want confidence and trust in what they're building the same way data engineers do on a day-to-day basis. Governance and quality make sure that we do data science puts things out.

Aaron Phethean (18:20)
Yeah

Exactly,

yeah, exactly.

Nik (18:29)
You have to make sure the

maths is right and if you sit there go, well, it's a bit of a black box, don't worry too much about it, they'll look at you and go, absolutely not, bye bye.

Aaron Phethean (18:35)
No way, I want you to trust that. Yeah.

It's a, it's really funny because, you know, everyone's focus is on like these agents. You know, I think of them as like, yeah, the apps, mobile apps, well, everyone's building one, never needed one. And, know, I might, but I am totally biased. Let's face it. I'm in a data company that we move data. We build a, you know, trust in the data. we're, very much. Well, I think is, you know, we have a tailwind of AI because our bit's going to be more important. You know, we're to be gathering data and helping people trust.

and organize it. Obviously the demand at the front end is only going to drive that even harder. And I think everyone in the data world kind of feels like that and a little bit of frustration going, but guys, the real work's back here where you're not looking.

Nik (19:20)
Yeah. And that's, yeah,

that's been a common problem forever in a day. It's been a common problem forever in a day. And like 95 % of people, 99 % of people's use cases are probably gonna, for at least for the next five years, gonna be a data warehousing problem. It's gonna be, give me my data set that tells me how I did yesterday. That's all people wanna know or show me where my problems are or do a bit of discovery around. It's going to remain the same.

Aaron Phethean (19:34)
Yep.

Nik (19:44)
And you've got really cool things like Databricks's Genie AI BI stuff that you can sit there, type things in and it the dashboard for you based on your data and things like that, which I think will be massive to this, just my interest in NLP. But I also think there's a bit of it, there's a bit of over-enthusiasm, like all things, when everyone sat there and went, big data will do everything. And then we went, okay, big data needs a big thing. And then cloud computing will do everything. And then this thing, this hype cycle continues on, and AI, agentic AI.

The with agentic AI is I always think of Hugo Weaving, like with his cool glasses from the 1990s. But like with agentic AI, it's going to be very specific towards things. And you have the CEO of Shopify, I think it was, turning around today saying, I demand to know why your job can't be done by AI or like why you can't use more AI. And it's like, well, mate, because most of the time you don't know what you're talking about because you're not a subject matter expert, but the people who do...

Aaron Phethean (20:14)
Yeah.

Yeah.

Nik (20:35)
haven't got the tools, all the education, all the understanding, and also they fear for their jobs. Until you can create a safe environment, the people are gonna be okay doing this stuff, they won't do it for you. So it is a tough one, but I think there's some cool stuff that will come up from the agentic world, but it's just how expensive will it be to do? And will it still be easier and cheaper for a normal human being to do it?

Aaron Phethean (20:40)
Hmm

Yeah.

Yeah.

And

I think that's the reality is that for a very long time, any job with enough effort could be automated away. but the capturing what you want that job to do, what you're the rules around it, excruciatingly difficult in a lot of circumstances. So one of the sort of possibilities of the agentic AI type world, it is a lot easier for someone to program it, for someone to control it, but they're going to control it really badly for a patch.

Nik (21:03)
Yeah.

Mm-hmm.

Yeah, there's a meme. Yeah,

but like there's a meme of the software engineers when AI started coming out. like, don't worry, boys. Like they've still got to figure out how to tell us exactly what they want because engineers still can't. Like they still can. I just want a dashboard, but you're automating a process. Yeah, but I want a dashboard. It's not automating a process. It's like, but I'm going to automate these processes. Or like you think about how many data nerds start to their teeth on, or whatever term it is, like on VBA.

Aaron Phethean (21:39)
Exactly.

Nik (21:47)
And that didn't do what we wanted it to do, but we made it and we had to force it and we have to use brute force. That's right. Like there's a reason that there's also like kind of critical attitudes sometimes taken. Some of my peers have got it where, oh, the kids who come out of uni now just throw computer things. And it's like, because that's what will happen. You'll have the end users who just sit there and go, yeah, I could do this and it's automated. Yeah, but it's costing a thousand pounds a week to run. And so like your salary times that by 16 and you've got hundreds of thousands of pounds worth of

Aaron Phethean (21:49)
Yeah.

Yeah.

Nik (22:15)
costs at the end of it, but people go, it gets hand waved away with AI at the moment. We say, don't worry about that. The ramifications will be immense. Okay. But I'm a CFO who's looking at that guy and expensive justify what you're up to. Yeah.

Aaron Phethean (22:20)
Yeah.

Yeah, that's exactly. And that does

get a touch on one of the sort of my pet interests. I like to think about the scale of people's problems. you know, if I'm talking to a company with a handful of employees and lots of widgets, it's very different to a co-op where there's just lots and lots of things.

And yeah, the, the, so again, the scale of processing that is kind of interesting to me. So how much data are we talking about sort of floating around your domain? What, what's, and how successful are they at the kind of warehousing of it?

Nik (23:00)
Lots,

lots. We're going through quite a drastic transformation at the moment anyway to kind of modernize a lot of our own stack, moving into much more serverless approaches, being able to accept streaming, IoT kind of things. So yeah, is an inordinate amount. I can't tell you exactly how much there is, but there is an inordinate amount of data. We're pretty good when it comes down to it. Across the whole organization, it will be petabytes because it's just that kind of org.

We're pretty good actually. We've got quite a hardcore data governance team who are really shit hot on the retention schedules. So we don't have ridiculous sums of data floating around inside a swamp. just like, if we've got stuff from 2009 still in there, actually they're pretty good with that. So we don't have huge swaths of data floating around, but what we do have is lots of data, very complicated data that we've had to link together.

trying to figure out who people were before we had integrated systems and like customer platforms and things that are being modernized whilst we're improving them. Yeah, it's an absolute nightmare, but also it's the bread and butter of what we're doing for us, what we're paid for. We're bloody good at it as well, if I'm frank and the team are incredible.

Aaron Phethean (23:53)
Mm-hmm.

Which is always hard. Yeah, it is really hard.

Yeah. I

think that that point about, you know, archiving off or deleting or, you know, the management of data that seems to be this sort of hype cycle, the trends of technologies, a little plug, I've got a LinkedIn live event on cloud cost in a couple of weeks. The point of bringing that up is that that reason I'm talking about that is that when cloud first arrived, there was this explosion of cost as everyone started to use it for everything.

And, I need some governance around it. Similar with your kind of cloud data warehouses, similar with all your SaaS tools and your kind of, it's really easy for these things to get out of control. And, you know, that, that discipline have been really good at managing it. It's one of the best ways to stay away from the cost.

Nik (24:35)
Yep.

There was a guy, there was a guy I

spoke, I spoke at data engineers, London November of last year, think in November, October last year it was, there was a guy Graham who was speaking about data redundancy. And like, all you have to do is think about, like, I always think about, I'm a dude who can talk in gifts and memes an awful lot. If I saved something on my phone and set it to four people, it's then saved on their phone, saved into their cloud. And that level of dissemination is a perfect, like silly iteration of this kind of process. Even in organizations we do that where.

Aaron Phethean (25:03)
Yeah.

Nik (25:11)
because we've got to be compliant with things. We've got to save it in S3 and ADLS Gen2. okay, why? because we've been told to. Right. Fine. Great. But then your cloud costs go up, but then you've got compute and it's, you doing the right thing? We're figuring out at the moment where like the perfect bandwidth for our current platform is like, and we're saving money by the fact that we're not having to compute loads and loads of things, or we're not rebuilding lots and lots of things. It's just Delta.

Aaron Phethean (25:18)
Yeah.

Yeah, yeah.

Nik (25:33)
Delta loads rather than just like full loads from old systems. it's, you got it. Like, and it's all stuff that's really, really good for data, but very often the costs that happen didn't need to happen in the first place if we were given the time to kind of design and build the things that were necessary at the time. But when does an org ever let you do that?

Aaron Phethean (25:35)
Yeah, incremental models.

Yeah, design well, engineer. I feel like there is

a trade off. I wonder, feels like the

one fine day, your data team sitting there happily building stuff and the CFO arrives and you guys are burning too much of whatever it is. Yeah. Whether it's ETL tool or warehouse. Yeah. Okay. The trade-off is that you could have engineered it a lot better. You could spend the time engineering it, but an incremental model might take like, you know, several man days, but only costs you several, several man days over a month. So you're well, how, how do we weigh that up? Do you dedicate people?

to it? How do you go about making that kind of sensible decision? you leave it to your team? What do you do?

Nik (26:28)
I think it's

about discovery. It's about just actually doing discovery properly. We're so keen as tech nerds to just stick our faces into it and go, we've got this cool thing that will do everything. And we kind of go, yeah, let's do it. Yeah, okay, great sledgehammer nut analogy. Like you don't need all singing or dancing. Sometimes a batch process just needs a batch process. Like, and it doesn't need all this kind of madness that comes with it.

That's something that has to kind of come culturally for that. That doesn't mean you can't innovate on things. You can innovate things. You can change, you can iterate things. But I think it all does start with the discovery process. So many data teams, so many people who I speak to as peers, as colleagues, as people who sit in the same industry comes down to the fact that discovery doesn't get done properly. And it's just a, it'll just be this data. Like just wedge it through. like, it's just, no one's given either the time.

all the respect to be able to discover the work that actually needs to be done properly. And some of it is just a, it's an API. Okay. Everyone's been using APIs for quite a while now, so we know how it works. But what does that model then need to look like underneath? What are the principles that we're going to be using to govern what that data actually looks like? And how are we going to put that together? Like we don't get time to do that very often. And the teams that do are then seen as being too slow. Where actually they're being methodical about how they do things.

Aaron Phethean (27:41)
Yeah, slowing down to speed up. Yeah.

Nik (27:41)
Um, so I think this, this, this trade-offs,

but I think it's a, cultural, um, pardon me, the bullshit cultural thing that came from American Silicon Valley. This move faster, break things actually doesn't work. It doesn't work because all it does is actually the compensation of that is burning people out, burning people out, giving people mental health problems, making them wake up and they walk into their office first thing or they walk into the home office first thing and go, shit, what we're going to see today, Right. That is far more damaging to a team and an individual than.

Aaron Phethean (28:03)
Yeah.

Nik (28:08)
actually just spending a couple of days making sure that your batch load works, your incremental loads done properly. But people aren't allowed to do that because you've got to add value and value is grand up until it all goes wrong because the central person who's been building that has gone off sick. So I think it's hard to sit there and say it's one specific thing, but it is quite a few of them, that's for sure.

Aaron Phethean (28:11)
Figure out what I think to do.

Well, you know, that feels like an absolutely awesome point to leave it on. By all means, innovate, but spend the time on discovery. Don't burn your team out. I mean, they feel like great, great things to live by. So thank you. You've been absolutely awesome.

Nik (28:40)
Yeah, yeah, it's easy things,

but hard things to talk about.

Aaron Phethean (28:44)
Indeed. Right. Thank you very much.