The Hidden Artist- Episode 14 with Cedric Wacker

Hi, Cedric. Thank you so much for taking the time out for this. I remember first coming across your LinkedIn profile, like, I think this was early, early November, I suppose. And then something about your profile just stood out to me. One was that you were a data nerd, and two, the fact that you were so open to having a conversation with people, especially since the fact that not everyone has that hunger and desire to learn so much about this world, without giving it away, because I feel like Only you can do justice to introducing yourself better than what I can. But that being said, I'm grateful for our conversation. And I remember the first time I asked you to be on a Zoom call, you came pretty prepared with a presentation to answer every question I had, which is not something which I've ever seen anyone do. So your passion, your energy, your desire to learn, and just like who you are as a person is definitely something which I'm really impressed and inspired by. And so without just rambling around, why don't you go on and introduce yourself and tell us more about who you are?

Sure, I'll introduce myself. I would actually love to hear the story how you found my LinkedIn profile in the first place. Like, it's such a needle in the haystack kind of situation. I was like, wow, okay, I did not expect that. And I've never received a message like that from someone who just wanted to hop on a call. But sure, so that in a second. My name is Cedric. I'm 27 years old, originally from Germany, have been making Copenhagen my home here for the last 7 years. I also received my education here at the Copenhagen Business School, where first I studied more generically business and sort of after my bachelor's figured out like, wait, so I'm not super— like, there was a lot of things I found interesting, but I was a little bit like, okay, what can I actually do? And by coincidence, I actually managed to land a job with a data company at the time, a consultancy, a small one. And then pursued a master, what in Danish we would call a, a, so like a IT and Business Administration master, where then heavily focused on data while having the, the opportunity to work part-time in that consultancy where I am a senior consultant today, and I've stayed with them for the last 3 and a half years-ish.

I love that. I'm curious to know what inspired you to move from Germany to Denmark?

To be honest, I grew up by the Danish border, so it wasn't like that outlandish. I mean, Europe is a pretty small place, and for us it was literally like, all right, I mean, the Danes are known for hot dogs, right? If we wanted to eat a hot dog, we would like hop in a car, 15 minutes, be in Denmark. So I feel like it wasn't that outlandish to make the move here. And actually, some of my close hometown friends also call Copenhagen their home today. So I feel like it's not a quite, quite unique story. I think a lot of us from, from the north, particularly Flensburg where I'm from, it's a small city, so you either go to Hamburg or Berlin or then orientate northwards, which would be Copenhagen, if you want to sort of get out of the small city and, and experience big city life and maybe pursue some opportunities that our hometown doesn't really offer.

But again, it's like even though I know that Europe is such a small country, if you say, but then it's still big enough because there's a language barrier that comes from learning German to switching to Danish, which is not easy. And obviously, when you mentioned about your passion for data versus business, which you originally plan on starting, so how do you— I'm curious to know, how do you go about— and you worked for a company before to learn that this was something which you were passionate about. So, you know, was that like, you know, what was that one specific moment when you realized that, hey, you know, this is what I'm more passionate about? And how do you go about, you know, learning more about that?

I suppose, like, is it correctly understood you're asking me how I discovered my passion for data?

Yes.

I think, so what happened there is that, I mean, Copenhagen Business School organizes lots and lots and lots of student events. And there was an event I actually wasn't aware of myself, but luckily some of my classmates were. And essentially we were done with lecture. Lectures, and it was an afternoon workshop on using Tableau as a data visualization tool. And that workshop was facilitated by the consultancy that I'm working for today. Um, and I just remember sitting there like, you know, I stumbled into that lecture hall randomly. I had not planned to go there, and I was just completely blown away by the charisma of the teachers, which would become my colleagues eventually, but also like just the field of itself. And then, I mean, it's fair to say that, you know, already like taking business administration in my bachelor, like the parts I enjoyed the most were certainly like forecasting and the like, right? So like I knew there was something, but I just didn't think that I had the talent or sort of passion or whatever it was. I enjoyed doing it academically, but that was kind of it. And then through that workshop, I learned that this workshop was leading up to a case competition that they were going to host together with Novo Nordisk, which of course now is like super well known for like Ozempic and Wegovy and whatnot. And certainly in Denmark, Novo Nordisk had been a huge player for years and years and years and years, decades. Right. But that was before all the Ozempic, Wegovy craze. So I participated in that case competition. I was lucky to have like super smart teammates as well and Yeah, we actually ended up winning the thing. So we got to present to like some SVPs, like senior vice presidents from Novo side, and then like the MD from the consultancy side. There was like a jury kind of setup, uh, presented our findings, they liked it. And then I could have gone left and take a job with Novo Nordisk, or right and take a job with the consultancy. And I think what always— so I think like I already knew from the workshop, like, hey, that's probably a place full of energetic, crazy smart people. I really want to work with, but then also the aspect of getting the chance to do multiple different things as in project-based work compared to zeroing in on one specific thing on the more corporate side of things was appealing to me. And I think still is appealing to me because I love getting the chance to, to learn about different businesses and areas of business, learn about their challenges with data. Which I find super valuable, particularly in the beginning of anyone's career, I think, rather than already specializing in, let's say, supply chain analytics or whatever department I would have ended up with.

Right. Got it. Wow. So what's your— okay, I'm curious to know, what's your favorite project that you worked on and any cool insights that you gained that you did not know about before?

I think you learn something every time. I really enjoy the project I am working for right now, or I'm working on right now. I can't disclose the client or anything like that, but it's a pretty big migration going from Alteryx, which is like a self-service analytics tool, to a DBT/Snowflake setup. And I primarily work as a data engineer/analytics engineer, right? So data engineering means you're concerned about designing, implementing, and maintaining systems that, you know, not collect as in the sensor itself, but certainly like get the data from the sensor into some sort of data warehouse and then do the data modeling and dimensional modeling and the like. So when you say data engineer, you're also, you're responsible for the integration of data, for example, using APIs. And then there's a new term which sort of was born as a DBT marketing term called analytics engineer. And really what that means is like everything what a data engineer does except the data integration, right? So right now I'm more of an analytics engineering project, which essentially means transforming data and implementing business logic because you get data in, in a very raw format usually. And you can imagine like, okay, there's some quality issues, particularly when it's human-generated data. Imagine somebody working with a supply chain system and putting in a number and then you hit the one, one too many times, or you have a comma instead of a dot as a decimal, or there could be a trillion different things that go wrong. So you want to go what the sort of modern standard or the current standard is really is that you have a raw layer in your data lake or lake house, which is the data in its raw state. And it could be semi-structured data or structured data. And then you have what you call a Medallion architecture on top So you have a bronze layer, silver layer, gold layer, sometimes a presentation layer on top, but that's often inside the gold layer. And really what that means is that we go from the raw data to we assign data types. We make sure that string values are cast as nulls if they're missing rather than just being empty fields, making sure that empty fields in numeric columns are zeros and the like. So that further downstream when we then start working with the data and relating the data to one another, different tables, or implement business logic, let's say, okay, we want to only consider orders that are in a certain step in the supply chain or whatever it is, that you then already have data that is standardized and treated and cleaned and the like so that you don't start building your models and then 3 steps down the line, you realize you have to implement a one-off fix, but you are moving in so many different avenues that you don't want to implement that one-time fix in 50 different places. So you rather implement it in the very beginning before you branch out and implement different business logics, sort of reusing the same table over and over and over again. Fine.

Wow, that was very informational. Yeah, and, and like, remind me, like, remind people who are watching this again, what time is it where you're at and what day is it today? Because you're talking to me about—

for me, for me it's, uh, 8:16 PM, and I'm a tad tired just because I, I had so much sleep last night. Maybe too much sleep, that could be the issue. But then Sundays are for long runs, and I feel like I really pushed myself today. For the rest of the day, I was in a very Sunday-ish mood, right, where I just kind of sleepy, which is great, it's a Sunday. But you know, now I'm I'm just trying to concentrate and zoom into this conversation to make sure that I'm not just yapping nonsense over here, right?

No, you're not yapping nonsense. The reason why I asked you is because you're saying this on a Sunday when it's late at night and you work the next day. But then that's what I mean when I say that you have that passion, right? You— I mean, and, and how open you are to have conversations and just to, you know, learn more about this. Space and just educate people what's happening in the world in general, and, you know, more about your space and how you do your work. And, you know, and I also have seen you write, which I will ask you later, but like, what's your favorite data visualization tool? Because I know you've written a few articles on that, um, on data visualization tools, right?

Yeah, mostly Tableau, which is a bit of a funny thing, right? Because I mainly work as a data engineer/Analytics engineer But I have an internal role that we call Tableau Tech Owner. And that's essentially a sales engineering role, just internally, meaning that I get dragged into different kind of meetings about people who ponder if they should migrate from Tableau Server to Tableau Cloud and the like. But then also I'm just trying to sort of make sure that we have the capabilities internally in terms of certifications and upskilling. And there's a new product feature, let's make sure all the colleagues hear about it and also know the use case. And sort of know when to recommend it for a given problem a client might face. So yeah, a bit of a weird mix, right? Because like mostly people do either front end or back end. And I think because I'm not a computer scientist or anything like that by education, I sort of started in the front end, which is lower barriers to entry, and then moved myself backwards into the back end, which is then also more pro-code heavy, right? Like mainly SQL, Python. So I'm like by no means a developer, but I know my way around SQL in Python pretty, pretty well. And Tableau and Power BI and all these kind of things, yeah, you can do scripting and you probably have to if you're like building a data product in there that is meant to actually create value in the business. But it's a bit more friendly to get into, which is also great. Like, so like personally, just because it was my experience, I would actually think if you're trying to get into data, that's probably one of the easiest avenues you can take, right? And then you can sort of see like, okay, where's my passion actually? Some people thrive on the more UX part of it, right? Which means like scoping with the clients, like what are the KPIs you're actually interested in, why, also challenging them on that, sort of conducting user research on who needs what, right? A management KPI is usually a lot more aggregated than an operational KPI. The manager wants to see almost a traffic light, right? Like green, yellow, red. Do I have to escalate? Can I just sort of leave the whole thing running? Someone operationally might want to dive into— can't say that word— dive into an outlier or dig into just on a way lower granularity, like try to explore the data. So that part sort of is the UX part, which would be part of a front-end project. But then also you probably would like, you know, sketch something in Figma or present it to the client and just sort of iterate a little bit before you actually start developing, because that takes a lot more time than iterating a couple of times with just like a wireframe. And then you start developing and then you have to like script in these tools and maybe you also need to do some data prep. And like by engaging in a project like that, I think you're pretty pretty soon going to figure out what is it that gives you energy, what is it that excites you, and where is it that you want to spend more time sort of deepening your knowledge. And if it's the scripting part and writing like the calculations and whatnot, hey, maybe that's an indication that something like an analytics engineer or data engineering or maybe even a data science sort of role is more for you, right?

So you started off, like, did you start off like doing everything or did you realize what exactly your niche was and you're most passionate about, which is data engineering and analytics, or was that something which you were assigned for and you knew exactly that this is what you were passionate about?

No, like, I'm just insanely lucky. And, uh, you know, I don't know how valuable that will be for anyone listening because I was insanely lucky to work in an organization whose philosophy is very much like we hire you because we believe you have the capability to grow and learn and the desire to. And we also don't want you to specialize early. We'd rather have you get on a solid generalist level and then deep dive into a specific area. Right. So in my day, I've been on Tableau projects, I've been on Power BI projects, I've been doing some Alteryx projects with this no coding whatsoever. But I've also done some data science projects, uh, and, you know, in the last 1.5, 2 years, lots and lots and lots of data engineering projects. Because then I just figured out, like, hey, I find this incredibly exciting. And, you know, like, I would literally—

like, if you see my YouTube feed, things I need to look at it now.

No, I'm not, uh, I'm not gonna screen share my YouTube feed, but, you know, like, you'd find your— you'd find occasional data platform, data science, data model. And I think like whenever you find yourself investing free time into a topic, or if you can send data engineering memes back and forth with your, with your colleagues and both of you get a good laugh out of it, then I think you're probably in a good place, right?

I think that's a great insight to be able to figure out, you know, what exactly you're passionate about. And I feel like, you know, like Okay, I had to, I had to stalk you a little bit to be able to come up with good questions and to be able to figure out how to, you know, what questions to ask you. But I feel like you learned all these coding languages, you know, while you were studying. And I feel like no wonder you call this luck, but I feel like you in a way, you know, by figuring out what you are passionate about, by seeing where you spend your most time and, you know, where you got a kick from. And, you know, I feel like that is something which can lead you to take on the projects that you feel most comfortable and at the same time passionate about.

For sure. But it's also, I think it's a tremendous amount of luck also, you know, like stumbling across a field or a job that you think is interesting. Like, when I was in my bachelor, I wasn't even aware necessarily that this was a job, that people would do this in their day-to-day. And if I knew the job title, I certainly didn't know what, what that meant in a day-to-day sort of setting, right? So I think I would encourage anyone really to just try lots and lots of different things, probably early, because there is a cutoff. I'm not a hippie, I'm not going to tell you like, hey, if you're 35, probably try 50 more things, right? But it's like, if you are in your 20s and you're listening to this and you're not really sure what it is you want to do, just try things and you might just get lucky enough to find something you're really passionate about. And for me, that was data.

I love that. That's great advice. I mean, I, I do believe in luck, and I, I agree, luck does play a huge role in this, but also so does your efforts into you know, taking forward that passion and what you, what by luck you stumbled upon. That kind of makes sense. That's great advice. Uh, but, uh, but that being said, I'm curious to know what does your day-to-day look like?

Um, so I'm going to talk about my day-to-day as a data engineer on a project maybe, and then day-to-day as a consultant, because those two are different things, right? Right now I'm on a quite long-running project. Um, I started early December. I'm definitely going to be there until like end of April, probably a lot longer than that. And we run a pretty big data product, which is essentially a supply chain analytics data product that helps this company to do short-term and mid-term supply chain planning. So that is, if you think about it, like sourcing of raw materials, It's the different production steps and then shipping out these products, but also stock level management on all affiliates worldwide of the company. So it's a pretty complex process actually. I'm not a supply chain person, so I probably got something wrong, but I was just in awe. And to be honest, information overload, how difficult it is to produce and ship something globally, right? It's like there's so many things that can go wrong, so many rules and regulations, so like, you know, just-in-time delivery and like you got to forecast demand correctly. And if there's a marketing campaign that could influence things and what are the sales cycles and yada yada yada, there's so much here that can go wrong. So it makes sense that people need good information to make good decisions on that. And we are a pretty good team, pretty— yeah, we're a great team, but also like a large team and we are globally. So half the team is sitting in India and then half the team is here in Denmark. And then there's different specializations. There is a front-end team, which is like Power BI and Power Apps developers. And then there's the back-end team, which is Snowflake and DBT in our setup. Snowflake being a data platform, you would call it these days. So you can store structured and unstructured data. It has storage and compute, but it also comes with lots and lots of additional functionality that a different traditional SQL server wouldn't give you. And dbt is a tool you can use for transforming data that plugs into Snowflake and uses Snowflake storage and compute. So it doesn't store data in itself, but it's super, super neat because it's really, really good for auto-generated documentation, tracking column-level lineage. So where does this column actually come from? What transformations have been applied to it until I get it in my hands. And the same for tables. And it's really good for what you call CI/CD, so continuous integration, continuous deployment, when you have like a large code base that a lot of people are working on and you gotta make sure the version control is there correctly and like everyone's contributing in a meaningful way. Day to day, how we ensure that there's technology, Azure DevOps, which is sort of like, you know, like making sure there's tickets and user stories and epics and like. We have DBT and we have a process around it, and it's actually agile by the book. A lot of people are going to tell you, oh, what we do is agile, and it's not really agile. This is actually agile in its purest form that I've ever seen. There's a, you know, dedicated product owner who's the business responsible for actually owning the solution. There's a dedicated Scrum Master whose only responsibility is making sure that we collaborate well as a team. And how we do that is that we have a daily standup in the morning, and then our sprints currently are 2 weeks long. And then we're going to have a review on our sprint, and we're going to have a retrospective where the retrospective is a bit more meta on what is going on and the review is more hands-on. And then we also have a sprint planning section where we sort of look at what is the overall goal of, of of our product here, what sort of bugs have been reported in the last couple, in the last 2 weeks essentially, and then they're allocated across the teams and the Scrum Master is collaborating with the product owner to sort of ensure good prioritization on these items and then we allocate them to the team members and then we run with it, right? So that is sort of, I like coming in early. Early for me means like 8 AM, so I probably have a good hour of focus time. 9, 9:30 usually stand up. And then depending if it's a problem or— yeah, what do you even call it? So it's supply chain. I'm not a supply chain person, right? So sometimes I can just code away because the business logic is clear. And sometimes I need to collaborate with someone from the business to actually ensure that the logics I'm implementing are aligned with the logics that the business wants, right? Because if the— if it says like logistic status 22 and I'm— I don't know what it is, right? So like, why are we interested in logistic status 22, right? Would be nice to know. And then sometimes it becomes a little bit more clearer what you're actually developing on this one.

But honestly, I, I get it. It's like, it's so complicated to be able to work on a supply chain project because it is so much going on in that space. And you know, one thing goes wrong and the whole chain is disrupted, which is, you know, it's a pretty big project. And, you know, you should definitely, you know, like taking on this project definitely is not easy, but the fact that you're doing it while you're in your early 20s is a big thing. So, you know, it's definitely something you should be proud of yourself. But I was wondering, you know, what's like one, you know, there's like so much happening in this world of data, right? There's so much happening. How do you, how do you keep up with these trends? And, you know, what is one go-to resource that you go to to keep yourself updated? As to what's happening?

I think the honest answer is you probably can't keep up, right? Like, it's impossible to be a specialist in every single area of data. So there's a reason why there's like analytics engineers, data engineers, data architects, data scientists, right? And I think like in consultancy, people like to talk about a T-shaped profile, which sort of means like, you know, something about a lot of things on a high level and then you have the leg going down in like one specific area, right? So I think like in data it pays, or probably what you should strive towards is having like a T-shaped profile. And I do it mainly like working in a consultancy. We are a premier partner with a lot of technologies like Snowflake, DBT, Microsoft Fabric, Databricks, and the like. And because these, these companies, like they often work with implementation partners, lots and lots of, lots of companies have super capable people, but we also get some additional expertise. And so these data companies know that, and most of them have partner programs. So for example, end of March, there's going to be a Snowflake partner meetup or user meetup, actually, that's not specific to partners. And then you go in for a day conference sort of type, right? And they're just going to demo some of their latest and most requested features. And then there is some practical hands-on sessions and you can talk to other Snowflake users. And then LinkedIn, I think if you follow the right people, definitely great tool. Medium has lots and lots of good articles as well. YouTube, I'm a massive YouTube fan, love YouTube. I spend way more time on YouTube than any other streaming platform whatsoever. So I think these are great. Great platforms and avenues to sort of keep up. But it's impossible to know everything what's happening in large language models if you want to consider this as part of data and know everything that is Microsoft releasing on Microsoft Fabric right now and know everything that Snowflake is doing. But a lot of these sort of things are also underlying currents that you will find in different shapes or forms across technologies, right? So like there's one big thing like OpenTable format. So like Apache Iceberg tables, for example. And then you're going to find it with 5 different names from 4 different providers, but like the underlying technology is all the same, right? So like you also got to distinguish between like marketing a little bit and what is the underlying, like what is the use case that's being solved with this? And I think if you focus on the use case, so like the actual problem that is being addressed, and then try to like sort of mentally map all these different terms to, to the problem. It's easier to learn problem by problem than trying to map all of these different phrases. And, and because that's marketing, right? They're all trying to do the same at the end of the day.

I love that. You can never keep up, and there's so much happening that, you know, you would bet, you know. But you're right, there's different names, but most of them doing the same thing. But now I'm really curious to see your YouTube feed. I really hope that I could see that point, because you said that there's so much— I mean, you rely on YouTube the most. So, you know, I can only imagine how you feel like given the videos that you watch and the videos that you get recommended because of the videos that you watch. But then you mentioned that you understand the key trends from, you know, the insights that you've gotten so far. So what is your take on— you know, this could be a broad question. You can take it whatever direction you want to take it into, but like like, what key insights do you see from all the projects that you worked on? And, you know, where do you think the world is heading, heading towards? Like, any investment trends or anything of that sort? Take us away.

Damn, now I have to ponder, right? Because it's very broad.

Whatever you, whatever you want to, wherever you see it.

I think just because, like, everyone's talking about AI, I think, yeah, that was one of my questions. I think it's kind of interesting. It's interesting to see the interplay. I think for me as a developer right now with AI, right? Because a lot of people might ask themselves, should I even get into data if AI replaces all of us soon?

Right? I was going to ask that. That was one of my questions. Can AI replace them? But you're getting in the right direction.

I think right now what I can see is that I'm definitely more productive using large language models, but they're a long way off from actually replacing you as a human being, mostly because you need so much interaction with the business. Because like, if, if like the end goal here in my case, right, my current project is to have a data product that helps the business make good decisions on their supply chain processes. The AI is very, very hard to know for the AI the semantics of the supply chain processes. Like if there's like, they have weekly meetings where like a supply chain guy sits with a finance guy, with a, you know, a marketing lady, whatever it is, and they discuss how they want to structure their processes. And now I'm over here and I'm supposed to build something for them. There's so much meaning and so much business need and so much things that I don't even know. So I need sort of a supply chain babysitter on my side to help me sort of build something meaningful. I'm very well versed in the technology. I can tell you what is the best way of like computationally, like like, how are we gonna solve this in an efficient manner? And like, what are different technologies we can use for it? But like, the actual business challenge I would never be able to solve without other ones, like another person's input. And I think trying to give all that information to an AI so the AI could do my job would be probably as much work as, you know, just doing it yourself in the first place. At least right now. And then the other interesting thing is like, yeah, the AI speeds up my work and this makes me more productive. But then data engineering, data in general, has a huge impact on the quality of large language models, of course. So they're all trained on a bunch of data, right? And a bunch of written literature and whatnot.

On.

But then in a business context, we often sort of want to, we want to tweak the models to our specific context. So like you get a model which is base trained from OpenAI or whatever, you don't need to train a large language, it would be too expensive anyway, but you don't train a large language model yourself, right? OpenAI or like Meta or whoever is going to be the provider is going to train the model for you. But then you take the model and if you have something like DeepSeq, it's actually computationally so efficient that if you have like like if you have enough compute internally, you can run it internally. So none of your data will ever leave your premises and circulate the internet. So now you suddenly can like train the models on more specific problems and insights and so on that relate to your company, right? But like how good the model will be in giving value back to your company entirely depends on the quality of information you feed it with, right? So like everyone wants a large language model, everyone wants advanced AI, but then the data foundation, if that's crap, your AI is going to be crap as well, right? And your model needs to live somewhere and it needs to be fed data somehow. And like not only once, we need to do it on a continuous basis in a robust manner and we need to monitor the data we put in and yada yada yada and all of this sort of It's not like we empty out a bucket of data and now we go, right? No, it's a pipeline. It's like a literal pipeline of a continuous flow of information and bits and bytes and data. And somebody needs to design that system for us to use AI. And I feel like data engineers, we are sort of the plumbers, right? We are the plumbers of data. Everyone wants to be a data scientist because that's sexy, but somebody has to be in the trench and dig and put the pipe down. And that's what data engineering is, right?

Those are honestly such great answers, and just like learning from you. Um, you know, I know you said that, you know, it's a heavy-duty question, and I just put that on you, which, you know, honestly, it was not fair to you. But then, that being said, everything that you said made absolute so much sense and, you know, helped me realize about— and just, you know, the people who are watching in general realize where the world is headed. And, you know, especially with right now AI replacing analysts at a hedge fund. That was something which I read about, and that was something which I feel like it's not very hard to predict that people feel like AI— people can lose their jobs as AI takes over. But like you said, it's going to take time, and it's not an easy process. So I feel like people, especially consultants and analysts, have their jobs pretty much safe because that is not easy to replicate.

Although probably within consulting, the trend is going to go from having lots— so most consultancies are kind of built like pyramid shapes, right, where you have lots and lots of analysts, senior analysts, and then the pyramid sort of gets tighter on the top, right? And I think the pyramid is probably going to require less of a base because fewer people can do more work sort of in terms of research and like data collection and sort of analysis quicker. But then I think where clients can do the same, so it's just going to be less of a demand for that. So I think like where we're headed from a consultancy industry perspective is probably deeper expertise in a particular field rather than sort of being the workhorse for the client, if that makes sense.

That does make sense. Uh, but then I was wondering, okay, um, I, um, I feel like there's so much bias when it comes to data. Like everything, you know, is heavily reliant on data, right? From business decisions to decisions being taken on, you know, giving loans or credit cards or just like, you know, every small little thing like a person gets a job, right? That's all dependent on data. But then that means that there's so much bias in this. So I was wondering, what's your take on how could, you know, this bias be eliminated to be able to make the best decisions while relying on data?

I think, you know, like, bias is, is a human thing. It's not a data thing. Like, the bias gets into the data if the data is based on human behavior, because we are all biased and making biased decisions. Or like, when we look at the data, whatever we read out of it, I don't think you'll ever get rid of bias, to be honest. You can try for the best, and I feel like a lot of companies do that. But also, I think there's actually a huge discrepancy between how data-driven companies think they are versus how, how data-driven they actually are. And I think in any business, if you were— like, I'm not part of boardrooms, right? But I wouldn't be surprised if, like, major decisions are mostly being made based on, like, the politics in that boardroom and sort of capital, like political capital that people have to spend rather than data. So I think actually the further to the top you go in a company, in most companies, the degree of data-drivenness will decrease. I think the sort of operationally involved folks are way more data-guided and driven in most places. Probably. Yeah, probably that's true. Way more driven and work more hands-on with data and actually use data to make informed decisions. And the higher up you go, I think the more political it becomes.

I, I think I, I mean, I agree with what you said on it's definitely not easy to get rid of bias unless, you know, even bringing on more diversity, they say, is going to solve that early. It's, it's, it's not easy to get rid of that, especially if politics is involved in that, or, you know, hands in the power of powerful people.

There's— but there's— if I can drop a book recommendation, there is actually a book called Calling Bullshit The Art of Skepticism in a Data-Driven World, and it's such an amazing book. And if you're just interested in, you know, how are clinical trials misrepresented, or how are selection criteria for universities, or advertisement, or just in the news— like, it's not all— it's not out of— and now I'm lacking the word, right? I'm not an English native speaker. It's not bad intent, always, right? It would be false to constitute that that's bad intent. Sometimes it is, but other times it's literally just that humans— like, the way that we don't— we have an inherent understanding of gravity because if I drop something, it falls to the ground. We don't for, like, the quantum world, right, in terms of physics. Like, we don't have an innate understanding of that. Or like the distance between the Earth and the I think the same is actually true for statistics in many ways. It's not something that comes intuitively to, to us, right? And you really got to study hard and then sometimes you're going to think that you wrapped your head around a concept and 5 months go by and you have to reread on it, to be honest. So I think, yeah, this whole book, it just encapsulates all that and more in a more elaborate way than I could ever pitch it. But, you know, if you have the time, read it.

I would absolutely love to read, and I'll make sure to link that, um, you know, Amazon link to that book in the description of this video so that more people can read about it. But then, you know, my whole understanding of the biases came from this Netflix documentary called Coded Biases, um, which I'm sure you must have watched it. If not, it's something, um, it's okay, you've read a book that solves that, that speaks more about, you know, the problem and solves it, etc. But, um, it's a documentary where it and it spoke about all these different biases that come from the data that's been put in from like especially like white people. And, you know, it discriminates more towards, you know, people who are of color, which is because in the past, you know, things have happened wherein these people, you know, it is very biased and gendered, like, you know, coming when it comes to refusing credit cards or just like loans and gendered or insurance to people of color. Just because of all these biases that have been, you know, just even when it comes to job applications and things like that, you know, not, not to specify more on my limited knowledge, but then that being said, there's biases in this part. What you said, you know, reading the book, that's definitely something which would help understand more about this and see, you know, what can you do to solve that.

One, one example that was floating around a few years back, and I, I read the headline, right, so don't quote me on it, but I think at some point there was an issue with the Apple Watches and other fitness watches that they had a hard time measuring the blood oxygen of people of color. And it's because like all the engineers who worked on developing it were white, right? And they tested it and it worked quite fine. But like apparently there was something because of the pigment they use, that green light to sort of— I think it reflects on oxygen particles in your blood. And like, I mean, you're a biologist, right?

Yeah, I did. I did read about it.

That's your area of expertise. But I believe it took some time for them to fix it, actually, and they hadn't accounted for that before releasing that. And that's probably, again, not because they were meaning to discriminate. It was just that they had a sample bias in the people they tested it on or who developed it. Right. But like for the company, it's just market share they're losing out on. Why would I, as a person of color, buy a device that doesn't work for me, right? So that wasn't the intent, but it probably was a product of sampling your, your test subjects or developers not in a representative manner, right?

No, I agree. Like, the very first way this documentary starts was a woman of color trying on a filter and it doesn't work just because she's of color and the data that it has been, uh, you know, trained on is of people who are white. And so that's why it couldn't just detect her face. and that just seems so interesting. But you're right, you know, I, I don't know if I read that paper or not. Seems very familiar. It talked about— but then, you know, after this, I'll definitely go back and read about it. But you're so right, it's like, you know, there's so much biases that just— and it's sometimes not even intentional, but then it just is there and it's inherent. And probably, you know, with like all the new rules that have been set up in America now with, you know, the hiring process I don't know how are they going to solve that problem. Good luck, America, I guess. But, uh, that being said, it's definitely something which is inherently present.

I think there's probably like— I'm, I'm a huge, huge defender of sort of equality of opportunities, not equality of outcome, right? I don't believe in like quotas or something like that, but I do think it's probably a good idea to hand in CVs without a name and without a picture, right? And just looking at the, at the qualification. Especially for these jobs where like thousands of people are applying to, and there's going to be like 3, 4 interviews anyway. So you're gonna have like plenty of times to sort of vet that human being in terms of, could I see that human being to be a great team member, right? Like we have this sort of test that we do when we're hiring where it's like, would I like to drink a beer with that person after work as well, right? Very Danish, very European. But like, it is also important for like building functional teams that there's of course the, the professional qualifications first, but then also like, will that person work in the team that we have today? But I do think for like bringing about a fairer society, probably cancel like names and, and pictures. That being said, I'm not a fan of like quotas or anything like that. I wanna— I do believe that most people are probably not racist and have good intentions and, and the like, and that's just what I would choose to believe, right? Um, it's one opinion. I could be right, I could be wrong, but, uh, that's sort of how I think about these things.

No, I think you're right. I mean, like, I feel like people deserve to be hired for their merit rather than for, you know, just being of a certain race, color, or gender or, you know, whatever, right? I mean, but I feel like, yeah, I mean, it all depends on, A, is this person qualified enough for the job? Is this person capable enough of doing this? And C, like he said, very Danish, would I want to have a beer with that person after work? But then it is so true, right? I mean, you all work on a project, you all work as a team, and it's not like a one-person thing, right? It's like everybody contributes to the team in general, which is why a person is there in the team at that organization in the same place. So instead of of looking at that person as a threat or just like another person, somebody who's there because of, you know, whoever that person is, is like, you know, that being said, it just plays a huge role in the dynamics of your colleagues and just like in general prevents the organization from achieving the goals that it's supposed to be achieving. So that's a valid point and I hear you and I feel what you said. You know, before I jump on to asking more questions, what What is it that you wish you knew before you entered this industry?

Honestly, I, you know, we are in an interview and I should give you a good answer, but I feel like it's been just like a discovery process for me. I think there's some obvious ones, like it's definitely when you work with data, right? Attention to detail is a huge thing. So if you don't enjoy sort of going down a rabbit hole and like doing that diligently, maybe it's not the industry for you. And that being said, there's nothing worldview-shattering I could reveal here where I was like, ah, if I had only known this thing, I would have been amazing at it from the get-go. I don't think that's how it works. It's probably, uh, you learn and you fail, and sometimes that has to be painful, as in you troubleshooted some code for 4 hours and then a colleague looks at it and can solve that problem in 2 minutes, and you bang your head against the wall, but you learn your lesson kind of thing. And that probably never stops. You, like, the frequency of that occurring is gonna go down, right? But like, there's always something to learn. But I just feel like it's a function of time and effort and doing it. So there is no magic pill that's gonna, gonna set you up for success. Be curious. Like, you have to, you have to want to learn, and you have to be curious and passionate. But then I do believe that anyone has the ability to do the job. Like, I don't think, uh, you know, like, lots of people might hear like data scientist or data engineer, and like words like engineer or scientist might be loaded, but honestly, like, there's no reason to gatekeep it's not rocket science. Most of the concepts are like concepts when you hear them, it's like, of course you would do it that way, right? Of course you would treat data as early as possible. So if you reuse it somewhere else, it's already treated. I don't have to implement it in 5 different places. That has a fancy name, but like the concept, it's kind of like, yeah, okay, that makes sense, right? So I think it's probably the case also in your field that there's a lot of fancy words for kind of concepts that inherently make sense, right? So don't be discouraged by that. Maybe that's my tip— don't be discouraged by fancy, fancy terms that sound complicated.

I think that's okay, advice, you know, being curious and just— I mean, like I said before, like, you know, when this— when on a Sunday at 8 PM— no, I mean, assuming it's like almost 9 PM now, um, at your place. But then David said I feel like if ever I text you, let's say on a Saturday or Sunday about this, you will respond. You're always there. You always, like from every post that you've put up, everything, you know, from like all the posts that you've put up, you know, every single article that you write, you just think your description on LinkedIn in general speaks volumes about how passionate, how curious you are to learn more and how just enthusiastic you are about, you know, building relationships and telling people making people aware about what's happening in this world, right? So I feel like just like from you and seeing you and how hungry, passionate, and just how open and curious you are, I feel like, you know, I can see that, you know, like wherever you are at right now is, you know, just a stepping stone to exactly where you want to be. And it's only onwards and upwards from here for you.

Uh, well, I certainly hope so. But, uh, yeah, I think you're right. If you're like, if you generally enjoy what you do, I think I think the rest is sort of a side product of that, right? And when I post on LinkedIn, that's mainly because it's sort of a soft requirement of my job. Funny enough, believe it or not, I cringe a little bit every time I put something up. But then again, there seem to be some people who appreciate me rambling about some new Tableau data modeling features. So maybe it's not all for nothing.

Every, every, every— okay, from like every time you write, it's like you say that, uh, like, feel free to reach out to me, uh, you know, would love to have coffee if you have any questions, you know, always open to have conversations. That's what your LinkedIn says. So, you know, I'm sure that, you know, this is not a part of your job requirement, this conversation, but you started doing it because you're passionate about it.

I don't know, I, I do enjoy it generally. I think it's interesting because I think there's there's something to learn from, from anyone, right? Uh, so there's definitely a lot of things I admire about you. For example, just deciding to, you know, go and start a little interview series or like do some little side projects that are also not in your study or job description, right? And I generally just, yeah, I like people. Like, I like connecting with people.

I love that. Um, and that definitely will take you a long way. But then shifting from data, uh, okay, I had this question for you. If today were the last day of your life, which I pray and hope you live until 120, but let's say if today were the last day of your life, what is it that you would want this world to remember? Or how would you want to be remembered?

I remember this. I, I don't think— I'm not a person for legacy. I would hope that my, my, my mom thinks I was a good son, and I hope that my friends, uh, would think that we had great and fun times together and that they find a— I don't know, that they're not too sad. I don't know, I never thought about this. I would just want my immediate family to be somewhat okay without me being here and my loved ones. I feel like, you know, like you have to pick your battles. Like, I'm not a person who donates to charity, but I will give the homeless person in front of my supermarket, like, buy them some food or something because I feel like like that's the immediate impact you can make. So in terms of legacy, I would hope that the immediate people around me will think of me fondly, you know.

I'm, I'm sure they will. And if, you know, and whoever's listening to this right now, and me for general, you know, I look up to you and I can say that, you know, this all is true because you live the life that you, you wanted. You, you know, may create an impact in this world and, you know, that's, that's all you want. And, you know, your loved ones love you and people who you had an impact on to remember you for who you are, which, you know, and I, and I pray that you live under 120 or for, you know, for like as long as you— as long as like for long enough. Um, but, um, I was curious, okay, what advice do you have for someone, um, who's feeling lost and doesn't know what next to do in life?

I think it's very easy to fall into a victim mentality, and I've seen it so many times around me that people sort of build resentment because they feel like they've been wronged. And every time I sort of encounter that attitude, I, I try to emphasize— I think empathy is important, but I'm also a little bit always like, yeah, but you know, is there really nothing you could do about it? Like You know, like whenever I have a bad day, I go out for a jog. And I don't always want to do it. But I know that I feel better afterwards. Because like, it's just biology, you're just gonna get your happy hormones, you know, that's how we're wired. And I feel like for a lot of people, structure and actually doing something meaningful with your time will make you feel 1000 times better in the medium to long term. And like, we all know that, you know, when we get caught up on Instagram Reels, or if you use TikTok or whatever, YouTube, we all know, like, sure, there's also educational content, but we all know inherently, like, if I'm like spending 20 minutes just swiping in a brain-dead manner, right, that's probably not a, not a productive use of my time. And it's okay to do it, like, the doses makes the poison. But I'm just saying, like, you can't lie to yourself. Like, you can try, but like, you can never fully lie to yourself. So like, if I'm telling me, oh, this was a good use of my time, you know better, right? Like, even if you're trying to convince yourself of something, you know, which is inherently not true. So I think like, you know, eat healthy, get your sleep, get your exercise. That's your baseline. That's things you do for yourself. All the, or school, you would say in Danish, like all the extra energy, your profits that you're having, that you can allocate away. Give that energy to other people, and I'm sure you feel great about it, right? But don't, don't compromise on the things that you do entirely for yourself, and then good things will happen. It's like the snowball effect, right? I don't know if you ever read Atomic Habits Yes. And I think what do they say? Like the atomic habits are the compound interest of self-improvement. And I think that's actually something I like.

Yeah.

That's something I subscribe to for sure. Like, not that I would have like 1,000 good habits and I'm a huge self-improvement guru. I just feel like have your house in order and good things will happen, you know? And like be nice. It doesn't take that much to, to be nice. I feel like so much of the conversation is like, oh, like the left says this and the right says this, and like you're a bigot and blah blah blah. Just be nice, you know. Like most people, it doesn't take much, and most people are probably nice. And like give them the benefit of the doubt and then you'll be okay.

That's great advice. I mean, I feel like I'm listening with all my ears. I'm just like taking and soaking it all in and being like, okay, I want to implement that. Whenever I'm having a bad day, you know, I, I go to the gym every day because that's what I need. I mean, I haven't been able to go for like— I mean, I usually try to go in at least 6 days a week, but then, you know, that being said, like going to the gym, working out, doing things for yourself, and, you know, giving away the extra profits that have things that you probably have time for and that'll make you happy. Again, you're right, it's like, you know, doing the snowball effect is a big thing again. But, you know, probably you really— Atomic Habits after this, because it definitely was a great read.

You know, I'm like, I'm a nerd about few things. One of them is biking, and I absolutely love it because I think the first or second chapter of that book is about the British race biking team and how they sucked for years until they got a new trainer, right, who was just looking for that 1% margin of improvement in every single thing they do, from equipment to training to sleep to nutrition. And I feel like that's, that's very inspiring. And I'm not saying I'm succeeding all of the time, but I feel like I've learned, and it gets easier over, over time to like recognize bad habits or patterns and then You know, once it came to mind, that's sort of like the first thing you need to do to, uh, to improve, right? I heard it once, like the 4 steps. I can't cover it together, but it's something like you are unknown unknown, like you're doing something bad, you're not aware of it, you're, you're aware of conscious unconscious mind. Yeah, something like that, right? Where you're like first you're not even aware of any bad behavior you have, then you're aware but it takes a lot of energy and time to adjust, then it takes a little bit less time and energy to adjust, and then eventually it's kind of easy.

Yes, you know, yeah, the unconscious, semi-conscious, conscious, and subconscious.

Yeah, probably that's right.

I think so, I think so. You know what, Right, so that's, that's, yeah, that's, that's definitely something which you're right. It's, it's a lot of things we do unconsciously without even realizing we're doing it, but then to be, you know, to get in a conscious state and like immediately stopping yourself from doing that. But like you said, biking is definitely— I remember reading that example from that book, but you know, biking definitely is super therapeutic. And you know, I, I haven't done it in a while, but I, I really enjoy it. That's how when I first got started my fitness journey, that's how I was able to lose out a lot of it because of biking, and I enjoy doing so much, but it is super therapeutic. Um, you know, two last questions, one of them being, uh, what— okay, if, you know, in an ideal world, what and where would this— what is it that you would want to be doing and where is it that you see yourself being at?

I think like on a like sort of practical level, like if I gave my, my sort of career and learning journey a little bit further, I would probably eventually move to more of a data architect role where, you know, if you've been in the trenches and work with enough systems long enough implementing as a data engineer, eventually, hopefully you have such a good idea of the landscape of available tools and the reoccurring problems that any organization faces that given a particular set of constraints, you can cobble together the best design of tools you want to implement or use together and also processes around it that you are actually able to like more sketch it out on a more strategic sort of level where you can say, okay, given we have this budget, we have this amount of data literacy in our organization, these are the stakeholders we needed to run, you know, we need the data live or close to live or whatever it is. Is that the set of requirement is that you're then able to, to give good advice and sort of lead a team of, of developers to also implement it. I think that would be super, super cool, but I'm, uh, I'm a lot of years away from that also, just because like there's so much to learn. And I think it's very much the happiness of the pursuit rather than the pursuit of happiness, right? So like I'm looking forward to the journey towards that, but I'm also gonna look forward to reaching that level. I think Jimmy Carr said it somewhere, uh, at some point. You know, the British comedian?

Part of them, yes.

But he said it, and I was like, hey, that's a beautiful play of words. B, if it makes sense, like, you have to enjoy the journey, right? Because like, the journey is 90% of the time we spend. So like, you better enjoy the journey. You'll be miserable otherwise.

Yeah, if you want to enjoy the journey, you just gotta keep waiting, waiting, and waiting for your goal to happen. And then you need to do all of this, be miserable, and then be so close to getting the treasure. And because you were craving so much and not enjoying the journey, then you will give up eventually. Because obviously, if you don't like the journey, if you don't enjoy it, then, you know, what's the point of it all? Exactly. But one last question. What is it that you wish I asked you but didn't?

I think you did a pretty good job, to be honest, of the conversation and interview. And I think that is actually a good thing. I can't think of anything. I think you did a good job.

Thank you so much, Cedric, for taking the time off for this. I really appreciate you staying up. You've work tomorrow, I know, but really appreciate your insights. And being someone who is nodding out while you were talking and will go back and re-listen to what you said so that I can read more about this. But that being said, I, you know, really feel like I learned a lot from this and from you in general. And I hope that whoever listens to this connects with you and has more conversations if ever they have any questions, which I'm assuming you'll be happy to chat more about.

Awesome. Yeah, thank you so much. It's been a fun experience. First for me, nobody ever asked me to pick my brain, so It's kind of fun. And then, yeah, I'm sure like eventually going to run into some data challenge in your life because of course you do. And then feel free to reach out.

All right. Thank you, Cedric. Appreciate you.