This is episode number 383 with Aspiring Data Scientist Sean Casey.
Kirill Eremenko: 00:12
Welcome to the SuperDataScience podcast. My name is Kirill Eremenko, a Data Science Coach and Lifestyle Entrepreneur. And each week, we bring inspiring people and ideas to help you build your successful career in data science. Thanks for being here today, and now, let’s make the complex simple.
Kirill Eremenko: 00:44
Welcome back to the SuperDataScience podcast everybody. Super excited to have you back here on the show. Today we’ve got a very special guest, Sean Casey calling us from Abu Dhabi, United Arab Emirates. Very interesting episode, it’s going to be extremely useful for those of you who are specifically starting out, starting on your journey in data science, just dipping the toes into the water. Sean shares his story of how he got into data science, how he got into this field several years ago and what a crazy rollercoaster it has taken him on. Or what a crazy rollercoaster his life has taken him on that has led him to be where he is now.
Kirill Eremenko: 01:30
He’s doing data science in the space of visualization for a large company in the United Arab Emirates. In today’s episode, we’ll talk about quite a few things. We’ll talk about DataScienceGO Virtual, so if you were there, you’ll be able to relate to Sean’s story very well and you’ll be able to cheer along as we’re discussing the things that happened, the people he met. We’ll talk about creativity in data science, the necessity or not necessity of a formal qualification in data science. You’ll hear Sean’s story. We’ll talk about visualization, an amazing book that you can read in the space of data visualization, why it’s important. We’ll talk about the data science community and Sean’s tip for asking for help and why that’s important.
Kirill Eremenko: 02:13
In a nutshell, this is going to be a great episode. If you need that boost of motivation, that inspiration to keep going forward and to become the best data scientist you can possibly be. So, without further ado, let’s get started, and I bring to you aspiring data scientist, Sean Casey.
Kirill Eremenko: 02:38
Welcome to SuperDataScience podcast, super excited to have you back here on the show everybody. Today we’ve got a super exciting guest joining us from Abu Dhabi, Sean Casey. Sean, welcome. How you going man?
Sean Casey: 02:48
Good Kirill, how you doing? Morning.
Kirill Eremenko: 02:49
Very good, very good.
Kirill Eremenko: 02:54
What’s the time for you? For me it’s 7:30, how about you?
Sean Casey: 02:56
Yeah, 10:30, so it’s getting to the hottest part of the day at the moment, but it’s Thursday so it’s the end of the working week here today.
Kirill Eremenko: 03:07
Awesome. Is it hot in Abu Dhabi?
Sean Casey: 03:10
Yeah. Yeah, it’s up to 45 later today. 45 Celsius, so I think that’s-
Kirill Eremenko: 03:15
45 Celsius? That’s [inaudible 00:03:17]. What is that in Fahrenheit?
Sean Casey: 03:19
I think it’s 115 or something. 113.
Kirill Eremenko: 03:22
115 degrees? 45 degrees, that’s crazy. How do you cope with that? That’s like, I can’t even imagine going outside in that temperature.
Sean Casey: 03:30
Yeah, you stay inside for as much as you can. Everywhere has ACs so you just try and avoid the heat as much as you can. It’s hot, for sure.
Kirill Eremenko: 03:46
So do you run to your car, how do you get … like, how?
Sean Casey: 03:50
You get to your car but you have to drive with your fingertips because the steering wheel’s so hot when you get in there first.
Kirill Eremenko: 03:55
No way.
Sean Casey: 03:55
You have to wait for it to cool down. Get the AC checked every six months to make sure it’s okay. But it’ll start cooling down again the end of September, middle of October.
Kirill Eremenko: 04:11
Do they have emergency services in case your AC breaks and they bring you a portable one?
Sean Casey: 04:17
They don’t. It’s a good shout though. There’s a business opportunity there.
Kirill Eremenko: 04:22
Okay, awesome. Well, Sean, really excited to have you on the podcast. Tell us quickly how we met. It was like the most random thing.
Sean Casey: 04:32
So yeah, it was three weeks ago at DataScienceGO Virtual. I think it was the second day. I had spent the previous night at the keynotes, at the presentations, in the expo center and then moved to the networking center. You get paired with somebody for three minutes and I met people from all corners of the world, all corners of the … or all ends of the data science journey, some really cool people. The second evening then, the first person I meet in the networking center is you and it’s 1 AM for me, I’m standing on the balcony. It just blew my mind. We had maybe a 20 second chat and then, I don’t know, I was on my mobile because our daughter was asleep inside and if I was chatting to people in the networking center on the balcony, she wouldn’t have had the best of nights sleeps, which wouldn’t help anyone.
Kirill Eremenko: 05:31
Yeah, I think it was so random. We got connected. I think you were also probably the first person I connected with on that day. I’m not sure exactly. But I remember we connected and then I was going to connect, click the … you get the button, connect, so we could stay in touch, and then I just wanted to make sure that it went through and I clicked the other tab and I think that’s why the connection broke. Like, that’s it. I clicked the wrong button. But luckily, once you click the connect button, you get each other’s details so you can stay and touch.
Sean Casey: 06:04
Yeah. That’s happened, there’s been a couple of people that I’ve been in touch with since. People who are at a similar point in the journey to myself, people who are brand new to it. And just a couple of messages in LinkedIn, a bit of support when people share posts and it’s …
Kirill Eremenko: 06:22
That’s awesome.
Sean Casey: 06:23
Yeah, it’s been cool. And also, the presenters, a couple of the presenters, the guys in Zeal, I spent the whole time in that area just having a one-on-one chat with them around data culture. Plus the access we had was incredible. Jason, Jason Koo had a really interesting talk on computer vision and I dropped in a question at the end of the chat, or at the end of the presentation, and Roberto put it up to him. And again connected with him on LinkedIn later on afterwards. And he was able to share the paper with me that he spoke about in response to my question around bias in computer vision models, and how physics is being introduced to machine learning models to help them understand that this might not be the most accurate picture, or the most accurate decision.
Kirill Eremenko: 07:29
Fantastic, yeah. That’s really cool. That’s really cool you could stay in touch. So, people have heard from me about this event, we were promoting it, it was a free event, DataScienceGO Virtual, and moreover, there was 2,500 people, so a lot of people listening to this would have been at the event and they can relate to. But for those who didn’t make it to the event, just in a few sentences, could you describe why did you sign up and what your experience was, just to encourage others maybe next time to check out DataScienceGO Virtual.
Sean Casey: 08:04
I signed up because I’ve been listening to the DataScienceGO real events. I’ve been listening to the presentations from those shows for the last three years and always wanted, God, I’d love to get over to San Diego or I’d love to get to LA to one of these events some time and this allowed me to be there. To be at the virtual events, so that was why I signed up.
Sean Casey: 08:31
And what I took away from it was just the encouragement and the opportunities for learning that are out there. Emily Robinson’s talk on the first evening just stood out for me. It was just that motivation, that encouragement that yeah, it’s a journey, you’re on a journey, you can be at different point on this. You don’t need to worry about the label or getting the label immediately, as long as you’re enjoying it in you’re on that journey it’s worth sticking with it, that’s for sure.
Kirill Eremenko: 09:09
Amazing. And did you do any of the workshops?
Sean Casey: 09:11
It was 1 AM. I had work the next morning so I didn’t hang around for the workshops. I’ve been meaning to look back at them but-
Kirill Eremenko: 09:20 No, totally understand. That’s huge that you made it to 1 AM. That’s kind of like the only challenge, is the timezones. We had people from 123 countries and making sure every timezone is satisfied is really hard. But apart from that, if you’ve got the commitment, that’s totally cool.
Sean Casey: 09:38
Yeah, and it’s all available online anyway to look back at and to read up. Jon Krohn just made his Google Colab book available for everyone to take. I just couldn’t believe that, it was … It was that learning, for you to be able to access it that easily was just phenomenal.
Kirill Eremenko: 10:00
Yeah, awesome, awesome. Fantastic. Speaking of journeys, tell us a bit about your journey, because I asked you to describe it to me and you sent me this huge email which I had so much fun reading. Tell us a bit about your journey.
Sean Casey: 10:17
Yeah, so my journey into data science, data analytics. I started off with a mathematics and computer science bachelor’s back in Ireland. So, I would have had a foundation in object oriented programing and just the logic and the good solid foundation in the mathematics. And I very randomly ended up moving to Abu Dhabi to teach mathematics and computer science. A random decision but one I was very fortunate to be able to make.
Sean Casey: 10:55
I arrived here in 2005, spent some time teaching, some time in school improvements and professional development. Did a Masters in Education at one stage. And I was kind of at a point where I wasn’t getting a whole lot of personal satisfaction out of what I was doing at work. It was great to see schools improving, it was great to see students access better learning experiences, but my own personal satisfaction of enjoyment, I guess, in my job was waning a little bit.
Sean Casey: 11:34
So, I looked into different areas of what I might go down next after I finished the MA. I looked into accountancy for a while, wasn’t for me. I looked at doing an MBA, again, wasn’t for me. I ended up going back to Java. It had been 10 years since I’d looked at Java, professionally anyway, eight years. So, I went back to Java, did a refresher course in Java and I got chatting to a good friend of mine, Gráinne Dollan, who lives in Dubai, works for IBM, she said, “Have you looked into data science yet?” Because we would have done a similar course in university, back in Ireland. She said, “Check out data science.”
Sean Casey: 12:23
I can’t remember, did she send me to one of your courses first or did she send me to the Microsoft Professional Academy? But there’s so much happened so quickly once I dipped my toe into it. I started just banging out courses for fun. I was driving to Dubai a lot at the time visiting schools. I’d have one of the Udemy courses or the edX courses playing on the phone hooked up to the speakers in the car. I wasn’t watching the, obviously wasn’t watching the videos, but I was just letting it soak in while I was driving. Just the buzz I got off it, being able to spend 10 minutes watching a video or listening to a video when I was in the car and going home and being able to code that out in a bit of a race against a video playing in the background. Just learning skills, techniques for a 10 minute investment.
Sean Casey: 13:24
With the Master in Education, I could have spent three hours reading a research paper and feel that at the end of it I was no better off than I was when I started. I get that it’s a different type of learning and you have to be able to arrive at your own balanced argument. To get to that argument, you need source of information. But for me, the return on the time I invested watching a course on Udemy or troubleshooting a problem on Stack Overflow. Just the return was incredible. And yeah, just really, really enjoyed the journey into data science.
Sean Casey: 14:08
I’m not trying to suggest that I’m anywhere near the end of the journey, but it’s a journey and I’m very much enjoying it.
Kirill Eremenko: 14:16
Why did you enjoy it? What do you enjoy the most?
Sean Casey: 14:19
I don’t know what it was like when you were in school, but in Ireland, in your mathematics classroom in secondary school, the answers were at the back of the book. So, your teacher would give you homework questions one to 10 and it could be on anything, but you knew opening the back page, when you did all your work, you knew opening the back page that the answer you had in your copy book was the same as answer in the back of the book. You just knew it. [inaudible 00:14:53] that sense of achievement that yes, you’ve done it right, or accomplishment, you’ve done it right and you flip to the back of the book and the answer’s there, as you expected.
Sean Casey: 15:03
I get the same sort of a feeling from analytics. You can spend 20 minutes cleaning a dataset or prepping a dataset or trying to work out a formula in Python or in DAX and you eventually get there, you get it to do what you wanted it to do and it’s just that accomplishment. That sense of, right, you’ve learned how to do something new and this is your validation of that learning.
Kirill Eremenko: 15:36
Mm-hmm (affirmative). Okay. Because I was thinking you were going to say the opposite. I thought you were going to say that in mathematics in school, you get the answer but in data science, it’s an open ended question. You don’t know the answer until you find it and different techniques might lead to different answers. How do you know that it’s the correct answer?
Sean Casey: 15:56
So sorry, when I’m talking about that sense of accomplishment, the data science work I do in terms of predictive stuff is minimal so far in my journey. It’s a lot of reporting is what I’ve been doing until now. I haven’t done a whole lot of modeling.
Kirill Eremenko: 16:18
Okay, so BI reporting.
Sean Casey: 16:19
Yeah, yeah, yeah, BI reporting.
Kirill Eremenko: 16:21
But still, even there, how do you know that you’ve got the correct answer? Because you could structure a dashboard in many different ways.
Sean Casey: 16:28
Yeah, you can, indeed. I suppose that it’s accessible to the people that are going to be using it. That it adds value to the users of the dashboard. So, if we’re creating one on academic results, we’ll try our best to sit with the people who are going to be using it to find out what they need. So, what do you need to dashboard to tell you, so then, if you’re looking to do a calculation in DAX, that there’s a rolling average of students and it displays the way you want it to display or the way that your end user wants to be able to extract the information from. Then it’s, yeah, then it’s the right answer in my head. It might not be, but it’s the right answer in terms of what the users wants.
Kirill Eremenko: 17:17
Okay, okay, gotcha. I guess it’s that satisfaction of delivering usefulness to the end user. But in addition, I find how it’s different to school, high school, uni math is that there is so rigorous. It’s so like, okay, very structured. There’s usually just one way or one optimal way to get to the right solution and you follow those steps. It’s just basically like mathematics. Like it’s a science. Whereas here, there’s an element of creativity. You can get a right answer but in several different ways and I think the satisfaction is even greater because you came up with your own way to get to that answer.
Sean Casey: 18:02
For sure. And, going back to your first part about it, it’s adding value is, if it’s making someone’s life a little bit easier by being able to access a dashboard to get the information they need as opposed to having to trawl through the analytics themselves to get there, it’ll hopefully make their roles a little bit easier.
Kirill Eremenko: 18:26
Okay. Yeah, absolutely. Helping other people make their roles a bit easier.
Kirill Eremenko: 18:33
This episode is brought to you by SuperDataScience, our online membership platform for learning data science at any level. We’ve got over 2,500 video tutorials, over 200 hours of content and 30 plus courses with new courses being added on average, once per month. So, all of that and more you get as part of your membership at SuperDataScience, so don’t hold off. Sign up today at www.SuperDataScience.com. Secure your membership and take your data science skills to the next level.
Kirill Eremenko: 19:09
So, tell us a bit about the rest of your journey or up to now. So, you said you started learning analytics through sitting in the car, listening, going back, revising. What else did you do? How else did you invest into your learning curve?
Sean Casey: 19:30
Yeah. I got to a point where I had developed a load of additional skills. I did your Python and R, your machine learning courses. I did a load of stuff in Data Camp and edX on dash boarding, on Tableau, on Power BI and so on. And I had got to a point where I thought, right, you’ve all these skills that are developing but you’ve acquired all these skills. It’s time to get some bit of a formal recognition of that learning if you’re going to take a step into an analytics role.
Sean Casey: 20:12
And I enrolled in a Masters in Data Science and Technology and I’ve kind of put it on the back burner for the moment, for a few different reasons which I’ll get onto in a second. But I approached the modules in the Masters, if the module was coming up on machine learning or on visualizations or on Java, I’d enroll in a MOOC, on an online course in Udemy or somewhere else, on Coursera to get the foundations and the skill that was coming up in the module for a fraction of the price. And just was able to approach the modules then with a solid foundation. And thankfully have been doing really well in them.
Kirill Eremenko: 21:07
So did the online course just before the module came up and so you came prepared to the module in the real world course?
Sean Casey: 21:17
Exactly. And as a consequence, probably didn’t learn as much as I could have from the Masters module if I came at it fresh. But it’s the collective of the two that informs your learning. So yeah, I’ve kind of put it on the back burner for a while. I think I’ve six modules completed. I’ve put it on the back burner for now because I started a new job 12 months ago. I had a baby daughter nearly two years ago so time’s not … it’s not as easy to dedicate your time to a full module at the moment. The online MOOCs are a lot easier to complete.
Kirill Eremenko: 22:01
I wanted to know, why did you see the need for formal recognition of your skills? I think it’ll be a very interesting useful question for a lot of people listening, because they might be asking themselves the same question. Are online courses enough or do I need a certificate from a real world university saying that I have these skills?
Sean Casey: 22:21
I thought I did. I thought that acquiring a certificate from a university would be what I’d need to make that transition from the type of role I was in to a more analytics role. And looking back at it, I probably didn’t need it. Don’t worry, it definitely helped me because it started opening conversations that yeah, I’m in the middle of doing this. But the skills that I’ve developed from the MOOCs and the online courses are, they’re the stuff that I uses day-to-day in my role and they’re far more accessible to people. They’re far more affordable. They’re far easier to commit to.
Sean Casey: 23:22
You’ll see posts on LinkedIn all the time about people saying, what’s more important? Is it more important to have on-the-job training, online learning through your MOOCs, enroll in a Masters. You’ll see some people saying that, stop posting … I saw a post a couple of weeks ago and some guy saying, “Stop posting these certificates of online MOOCs, what you should be doing is working on Kaggle projects.” And I totally disagreed with that. I think it’s, you do what you enjoy and if you learn best through an online MOOC and you feel like that you want to accumulate a bunch of skills in this area before you can even think about starting a Kaggle project, or maybe a Kaggle project just doesn’t seem as the best return for you, I don’t think anyone should say stop completing these courses. Because it’s all learning. It’s all someone just trying to learn more about the area and trying to develop a set of skills in the area.
Kirill Eremenko: 24:25
Okay, thank you. That’s very insightful. Let’s talk a bit about the way … so, in addition to your learning, you told me before that you read Cole’s book. Cole Knaflic’s book about visualization in two days. I think I have the book here. One sec, I’ll just grab it. Actually, I have both her books right here. I just bought them myself a few weeks ago, so that’s book number one, Storytelling Data. She’s got a second one, I messaged her, I invited her to the podcast and she’s like, “So, which book are you reading? The first one?” It was like, “oh, you have a second book?” And then there’s a second one, it’s called Let’s Practice.
Sean Casey: 25:06
Need to get my hands on that.
Kirill Eremenko: 25:08
Yeah, hands on. So, I’m totally loving it. It’s called Storytelling With Data by Cole Nussbaumer Knaflic. Fantastic book. You said you read it in like, two days, in Thailand.
Sean Casey: 25:19
It was Vietnam.
Kirill Eremenko: 25:21
Yeah, Vietnam.
Sean Casey: 25:23
One of your guests, at the end of the show, you asked, recommend a book. And one of your guests recommended the first one, Storytelling With Data and the recommendation was so strong that I went away home that evening, bought the book, it arrived before myself and my wife went on our Christmas holidays, I think. Yeah, our Christmas holidays to Vietnam and I was sat in this lovely little café in Hội An, read the book in two afternoons and it was just that penny drop moment. It wasn’t that I was learning anything completely mind blowing, it was just stating the very obvious facts that you should have known when you were creating visuals in Excel or in Power BI or so on.
Sean Casey: 26:15
So, up until that point, I would have had got a dataset for a school I was working with, pumped it into Excel, ran off a couple of visuals and the visual was the last part of the step, up to that point, the visual was the last part of the step. So, you did what ever transformations you had to, and you produced a visual in Excel and you printed it or emailed it or whatever. But you never did anything to the visual. Whatever Excel recommend, you took their recommendation. After reading that book, the visual is only halfway along the process, because then you’ve got the formatting power to tell your story through the visual. So, simple things like just getting rid of noise, things that should have been very obvious to me before that point but you just needed to read it to realize it.
Sean Casey: 27:15
And playing with color, Cole is a big fan of grays and blues and it just runs throughout the book. I’ve tried to use that in many instances in my professional life just what is … A lot of my work would be around school inspections and you can create a visual in whatever tool you use and you can give it to someone and hope that they take the message that the visual is trying to portray. Or you can emphasize that message to a point that it’s impossible for the reader not to take a message. So, putting the noise to the background in grays and just emphasizing the key point. So, that’s been really powerful for me and the book just opened up my eyes to a whole new aspect of, a whole new corner of data science.
Sean Casey: 28:14
Up to that point I guess I had seen data science as Python or machine learning and that was the data science journey, onto deep learning, AI and so on. But this opened up a corner of it for me that there’s a science behind the presentation of information as well. And like you’ve mentioned already, it’s that crossover then between creativity and how you present that information is really insightful.
Sean Casey: 28:47
I got back from Vietnam. I think I’d already taken a Power BI course before that through Microsoft, but Power BI had changed so much since then. You’ve had guests on your show, Tableau had come up a couple of times on your show, so got chatting to a friend, said, “Here, what’s this Tableau thing about?” Same friend I mentioned earlier, Gráinne. She said check it out and you’ve got a free trial version with it. Played around with Tableau, it blew my mind man. It was just how quick it was to get really insightful visuals, interactive visuals that displayed a ton of information and used a ton of data in them. So yeah, that was mind blowing. And I used Tableau quite a bit in my work when I could but my role wasn’t, at the time, wasn’t solely on data. I had a lot of other hats I had to wear at the time.
Sean Casey: 29:57
So, I could see opportunities in analytics for me and yeah, that’s probably the next question. You can cut it there but there’s probably another question you’re going to ask in a minute about how I got [crosstalk 00:30:13].
Kirill Eremenko: 30:12
No, no, please keep going.
Sean Casey: 30:17
Yeah, sorry. So yeah, I think it was around, yeah, January 2018 I had been, I had a lot of modules done at this stage, a lot of courses done, a lot of new skills that I didn’t have a few years previously and I got a random text message from a good buddy of mine, Andrew, saying, “Would you like to go caddying this weekend?” And I was their-
Kirill Eremenko: 30:46
What is caddying?
Sean Casey: 30:48
Caddying is carrying someone else’s golf bag around a golf course.
Kirill Eremenko: 30:52
Oh wow, okay.
Sean Casey: 30:53
So yeah, so I’d never done it before but it was an invitational that was on here in Abu Dhabi, so there was a load of football players, ex Man United football players, like there was Peter Schmeichel and Dwight Yorke, who would have been the people we were roaring at the TV at in back in the end of the ’90s. Who else was there? There was Luís Figo, Alessandro Del Piero. There was a load more, Ruud Gullit.
Kirill Eremenko: 31:21
So, they all came in to play golf in Abu Dhabi?
Sean Casey: 31:23
They all came to play golf in Abu Dhabi and we showed up as part of a group to caddy for them. It was an invitational that was actually sponsored by the organization I work for now, GEMS Education. So, the two sons of the owner of the organization Jay and Dino Varkey were playing in the competition as well as a number of others. I got put on Jay’s bag. Jay Varkey’s bag, so I was, carried Jay’s bag around the course, had a bit of a chat with him. He said-
Kirill Eremenko: 32:07
You were probably hoping for a football player.
Sean Casey: 32:10
I probably was but you know, it probably worked out a lot better for me. I was probably hoping for Peter Schmeichel but I think it worked out a lot better for me. I got chatting to Jay throughout the round. He asked me what I did. I obviously knew who he was. It would be kind of hard not to know who he was over here.
Kirill Eremenko: 32:27
Even though you weren’t working in the company, you knew who he was?
Sean Casey: 32:30
Of course, yeah. GEMS, they’re-
Kirill Eremenko: 32:32
So, it’s a big company?
Sean Casey: 32:34
Big company, yeah. Very big in the UE. So, I got chatting to him, asked me what I did, I said, “I work in school improvement but I’m trying to branch into analytics, data science,” had a bit of a chat. At the end of the day, he said, “Look, if you ever fancy coming to work for GEMS send me your CV,” which was very nice of him to say. He didn’t have to say it at all, but very nice of him to say it at the end of the round. And then a couple of things happened in my personal life. My wife had told me the week before that we were expecting our first baby, so-
Kirill Eremenko: 33:15
Amazing.
Sean Casey: 33:16
Yeah. Incredible, incredible news and changes your focus. But then the following week, the company I was working for were going through some challenges and hit us with a significant pay cut overnight. So, I-
Kirill Eremenko: 33:34
Must be tough knowing that you’re expecting a baby to face a pay cut at the same time?
Sean Casey: 33:40
Yeah, yeah. It was probably the fire I needed to get moving. So, I said, do you know what? Jay told me to send him my CV, sent him my CV and Jay set up a conversation with my now boss Hywel Benbow who is the GEMS VP for data, global data and analytics, so I set up a chat with Hywel. We had a chat in a coffee shop in Abu Dhabi for nearly two hours one afternoon and called me for an interview. Went for the interview, thought it went pretty well. But there was some challenges around onboarding straight away. There was some … I couldn’t join immediately, so I took a different analytics job with the local Ministry of Education, stayed there for a year but always had my sights on the GEMS role.
Sean Casey: 34:44
I thoroughly enjoyed the conversation that I had with Hywel and the subsequent interview and could see that it was a place that I’d be able to grow, I guess, be able to grow in, grow professionally while also adding value. And then I was lucky enough to be able to join them last August and it’s been a lot of fun since. It’s been a lot of fun.
Sean Casey: 35:14
I think I said to you at the end of my email that I know it’s a journey. I’m never going to know everything in analytics. I’m never going to know everything in data science, but I enjoy what I do. I enjoy getting up every morning, going, all right, not going to work in the current environment. Going to different parts of the apartment. It doesn’t feel like work when you enjoy it. Sitting at the computer all day just playing around with data is very enjoyable and trying to manipulate it so the dashboard works the way you want it to. Or you’re doing some modeling that you’re trying to increase the accuracy as much as you can. It’s a lot of fun.
Sean Casey: 35:59
So, I’ve been very fortunate with just answering the phone call to my buddy that day, to getting an offer to send my CV if I ever wanted to join their organization, to being able to have a cup of coffee with my current boss. I’ve been very fortunate to get those opportunities but I’m eternally grateful to all the people who have helped me along the way in my journey.
Sean Casey: 36:29
I think the beauty about data science for me personally is that the community is so willing to help. It’s so willing, people are so willing to give you a little bit advice on the way or try and help you solve a problem or direct you to a different course or a different piece of learning. They don’t have to. They’re busy themselves. They’ve got their own demands at work and their own pressures in their personal life but people are still on, you can post a question on any one of the communities and you’re pretty sure you’ll have an answer within 24 hours. For the ones I’ve used anyway, the Power BI community or the Enterprise DNA community. There’s always someone there to say, “Have you tried this?” So, that’s part of the reason I want to continue on with this, continue on this journey.
Sean Casey: 37:31
Number one, I enjoy it. I enjoy it immensely. But it’s the opportunities to learn, or I’m never going to be bored or stuck for something to learn in the future anyway, that’s for sure.
Kirill Eremenko: 37:42
That’s awesome. That’s awesome. And you’re right. It’s important to enjoy what you’re doing and I think we’re all fortunate in data science that the community’s so amazing. It makes it easier to enjoy what you’re doing. Imagine if there was a very back stabbing careerist type of culture where you couldn’t trust anybody, nobody was willing to help. It would be quite hard to enjoy what you’re doing faced with that every day. So, I’m also very grateful for that.
Sean Casey: 38:13
I don’t think the area would be what is if it had that sort of culture that you just described. I don’t think the advancements which have happened so fast in the last five years, what’s happened so quickly, wouldn’t have been possible if there wasn’t that collaborative nature and the willingness to help and the willingness to share my piece of work. I go back to what Jon Krohn at DataScienceGO Virtual, he didn’t need to put up his Google Colab book for everyone else to take. He’d spent time working on that, spent time producing it, but he’s willing to share it. I think that’s phenomenal. You don’t get that everywhere. You don’t get that in every industry. And it’s because of that willingness to share and the willingness to put your work out there that the community’s able to grow and advance at the speed at which it has.
Kirill Eremenko: 39:16
Yeah. It’s absolutely fantastic. You mentioned there is some luck in your story by picking up the phone and going, being put on the right bag of the right person while caddying. Also, there was help from the community, which is amazing. But I think it’s important to also be fair to you that you’ve did a lot on this journey to make it happen. And with that, I wanted to ask you, what would you say is the one biggest thing that looking back or ability or skill or habit that helped you in this journey? Something that you can share and other people listening to this can replicate in their own journeys.
Sean Casey: 40:05
A hard one man. I think asking for help. I’ll go back to the asking for help when you need it is an important one. You will encounter challenges along the way. There will be hurdles that you’re not able to overcome or parts of code that you’re not quite able to figure out. But asking for help along the way, be it whatever, it doesn’t have to be an analytics journey. Whatever journey you’re on, asking for help when you don’t quite get something or when you just can’t quite hack what you’re trying to do or totally digest what you’re trying to learn, asking for help is a really important one. Because people are good. People are really good people. [inaudible 00:40:52] generally and they’re very willing to help.
Kirill Eremenko: 40:55
Why would you say that was a hurdle you had to overcome?
Sean Casey: 41:00
I guess it’s about your own belief in yourself that you might be able to do this on your own without-
Kirill Eremenko: 41:12
Like asking for help means you’ve failed, type of thing?
Sean Casey: 41:17
Yeah. That might be a subconscious thought in your head, but I think throwing that off early, no matter where you are in your learning journey in whatever area you’re learning in, I think that’s, throwing that off quickly and knowing that it’s okay to ask for help.
Kirill Eremenko: 41:37
Okay. How do you ask for help? Where do you ask for help?
Sean Casey: 41:40
My team is, the team I work with is incredible at the moment. And I think lockdown or remote working has really helped us with that. We’re a small team, but my boss Hywel will set up a time where we can go onto Microsoft Teams call and he’ll share a piece of his work from the last couple of days or I’ll share a dashboard that I’ve been working on. And you put your hand up straight away. I’ve hit a problem here. Can anyone here have a look at this? So, the team together will try and troubleshoot the problem on the screen. But that could be a first one if I’m at home trying to figure something out. By night, I’ll go to YouTube straight away because if I’ve ran into the problem at my stage of the journey, someone else has encountered it before.
Sean Casey: 42:35
Last night, my issue was around a refresh in Power Query taking incredibly long in relation to the size of the dataset I was working on and a quick video from, I don’t know if you know those guys from Guy In A Cube, it’s five minute videos on how to figure out your own challenges in Power BI. So yeah, I was basically putting too many marges into Power Query that I didn’t necessarily need. So, that was slowing me down. I go to the communities. There’s always someone in one of the communities who’ll offer help.
Kirill Eremenko: 43:16
What communities?
Sean Casey: 43:18
Power BI. Most of my work’s in Power BI, so the Power BI community, I’ll go there. I’ll go to the Alteryx community and someone will have published their workflow on the Alteryx community which you can just download and adapt for your own problem or your own project you’re working on. Stack Overflow if I’m working in Power BI or in Python, Stack Overflow’s definitely my go to if it’s an issue in Python. Unless you’re at the very edges of the data science space, someone else has encountered these problems before. They’re quick fixes. The code will be there for you to copy and paste and use in your own projects, in your own work. I think it just goes back to that collaboration and that willingness for people to share their work, put their work out there and let others learn from it and then take it further. That’s how it grows. That’s how we’ve got tech to the mind blowing space that it is in the last 50 years. It’s incredible.
Kirill Eremenko: 44:32
Yeah, yeah. Absolutely, absolutely. Yeah, so interesting. Your advice about asking for help goes back to not just, because first I understood as an external asking for help. But it’s a combination of asking for help externally and searching for the right answers that others have maybe already asked for and they exist. Ultimately, it is what you said in terms of being able to be honest with yourself and be kind to yourself that, hey, I don’t know everything. It started fine, I’ve tried to figure this out. Let me go check what others suggest. And so not being stubborn, I guess, and trying to prove to yourself that yes, I have to do it myself.
Sean Casey: 45:19
That’s an internal thing. That’s something that you need to … I’m not saying it’s something, it’s a challenge that everyone has but-
Kirill Eremenko: 45:27
That’s true. Success is 80% psychology and 20% mechanics. Most of the time what is stopping us from progressing in our careers is internal. So, it’s a very useful piece of advice that you’re sharing, that there are people out there who are probably stuck because of some internal stubbornness or fear of being an imposter. Or fear of being, feeling that they’re not good enough or that they fail. Fear of failure. And that is really preventing them. So, looking within will always yield much more progress than looking without.
Sean Casey: 46:07
Yeah. And just something you said there about feeling that you don’t belong. I mentioned this in my email to you but I remember that first certificate I got from one of your courses on Udemy, and I think it was the Python A to Z course. I’d seen loads of them on LinkedIn, I’d seen loads of other learners posting them on LinkedIn up to that point. And I got that first certificate. I can’t remember the exact day, so I’m going to guess it was some time around late 2016, might have been late 2017, I can’t remember exactly.
Sean Casey: 46:49
But I posted that certificate on LinkedIn, at the time I might have had 150 connections on LinkedIn. I wasn’t very active on it at all. But because I tagged yourself, SuperDataScience, people started seeing it. People from all corners of the world started clicking on it, writing a little encouraging post. It was like their way of saying, “Hi, you’re dipping your toe into data science? We welcome you. We welcome you with open arms.” It was powerful. Not that you’re doing it for the likes or you’re doing it for other people, that’s not why you’re doing it, but it was just the sense that right, the community’s happy to see someone else here and you’re not an imposter. You’re learning like the rest of us. We all have to learn somewhere.
Kirill Eremenko: 47:40
I love it. You’re not an imposter, you’re learning. That should be the tagline of this episode. I love it. That’s awesome. Sean, what’s next? What’s next for you?
Sean Casey: 47:48
What’s next? I enrolled in your data associate bundle and there was … that was free a couple of days ago. I think the whole team enrolled in it so I want to complete that.
Kirill Eremenko: 48:04
Awesome.
Sean Casey: 48:04
And start ticking off a few courses. I’d want to be able to spend a bit of time looking into computer vision and NLP a little bit more, but I’ve a few other areas I need to tidy up on first before I get there. Yeah, just keep learning man. Just keep enjoying this and keep trying to find better ways of doing what I’m doing already. I’m learning a lot in, every day, just on the job I’m learning a lot in the backend of Power BI and the Power Query part of it and trying to make, try to spend more time in there. And spend less time on the canvas if you know what I mean. Just setting it up right in there.
Sean Casey: 49:00
What else? Yeah, just keep having fun man, keep enjoying it. Keep sharing my learning with other people if they ask. Along the way, I’ve had a lot of people ask about … I don’t try and portray that I’m a data scientist by any stretch of the imagination, it’s a goal that I’d like to get to at some stage. I use a little bit of modeling every now and again but that’s the … But if people ask you, “How did you get into this, what were you doing?” I’ll always send them in the direction of a few different courses. At work, a lot of people ask about Power BI. They see the product of our work in the dashboards we publish and they’ll ask me, “Okay, where can I start learning?” We’ve got the licenses to share with them and it’s share learning opportunities, share courses, just let people, welcome people. I was welcomed in, welcome other people in.
Kirill Eremenko: 50:06
Fantastic. Well, very inspiring advice. Sean, this slowly bring us to end the podcast, I wanted to ask you, to finish off, what’s your one message to those learning data science? Those that are starting out this journey, people who are feeling just like as you were, dipping their toes into this field. What would your one biggest piece of advice be for them right now?
Sean Casey: 50:36
To start small and all of a sudden new aspects open up very quickly. When I say start small, take an online course in a data vis tool or in a programing language and once you’ve completed it and you still like it, all of a sudden a whole new set of doors open. And when I say doors, I mean doors within that learning journey. So, I had no idea when I started out in data science that I was going to end up spending most of my time in Power BI. That was a door that appeared after I’d learnt a certain amount of skills already, or developed a certain amount of skills already.
Sean Casey: 51:29
And that’s another part of it too with the learning thing is, there was a challenge recently … Yeah, so there was something I hit recently on using a rolling average in Power BI. It’s the same in using it in, hitting a problem in another area of a programing language. When you learn how to do something differently, you then start applying that new learning to your work, to your, be it your dashboards or your code or whatever. Until you hit another new problem because of what you learned with this problem. I’ll just take an example, all of a sudden I can do rolling averages. Now, the next part I’m going to hit is I’m going to hit a challenge around rolling averages that are split over different quadrants or different … So it’s [crosstalk 00:52:26].
Kirill Eremenko: 52:26
I think it’s called a threshold concept. Because once you learn it, it’s something you can’t unlearn and makes you see the world differently. Now that you know rolling averages, you’re always going to think, oh, can I apply rolling average here. You’re always going to see those same things that you saw a year ago but absolutely differently because there’s potential for you to apply this new skill.
Sean Casey: 52:50
Yeah, definitely. And until you hit the next problem, and then you’re better. You hit the next problem, you go away, you learn how to solve it, you ask for help, you apply it and you’ll hit another problem again. We’re never going to be bored anyway, that’s for sure.
Kirill Eremenko: 53:06
So, basically, start small and if you like it, progress in that direction. If you don’t like it, try something else.
Sean Casey: 53:12
Exactly. Because there is so much to it. There’s so much in the data science, data analytics area. You don’t have to be working on the same tools. The tools are adapting and being produced and being released quicker than we can keep pace with. But it’s the skills. It’s the way you approach it, it’s your thinking that will get you through.
Kirill Eremenko: 53:43
Awesome, awesome, thanks Sean. Great advice. Great advice. On that note, we’re coming to an end. To wrap up, I want to say thank you for coming on the show. And also, before we finish off, before I let you go, where’s the best place for people to get in touch with you? Maybe they have follow-up questions, just want to connect, network with you.
Sean Casey: 54:05
So, LinkedIn’s the easiest one. Sean Casey on LinkedIn. I’ve taken a bit of inspiration from Emily’s talk at DataScienceGO Virtual and I’m in the middle of hopefully setting up a blog post as well. It’s not there yet but it will be and once I have that ready I’ll let you know because as Emily said in her presentation, and as I’ve heard from loads of other people already, that you have this knowledge now, don’t keep it. Share it. Let other people learn from it. So yeah, I’ll have a blog post later on. It will be on InsightAndAnalytics.com, but it’s just not there yet. It might be by the time the podcast airs.
Kirill Eremenko: 54:52
Maybe, yeah. If you put in a bit of work very soon it might go there soon, it might be there when the podcast goes out. Okay, fantastic. And so, LinkedIn and you said Insight and Analytics?
Sean Casey: 55:08
Yeah, InsightAndAnalytics.com.
Kirill Eremenko: 55:12
InsightAndAnalytics.com. Awesome. Well, fantastic. One final question for you, what’s a book that you can recommend to our audience?
Sean Casey: 55:18
I think you probably have it within reach there, do you?
Kirill Eremenko: 55:22
Ah yeah, this one. Storytelling With Data. Definitely.
Sean Casey: 55:25
Amazing book. That doesn’t have to be for people that work solely with data. Anyone that presents information in any aspect of their role, if you want to make sure your message, the message you want the audience to take from the visual is what they take, that book’s definitely going to help you.
Kirill Eremenko: 55:47
Fantastic. And it’s such an easy read. It’s big because, like as in the size, the height and the width is big because the images, but there’s a lot of images and there’s a lot of margins. I can tell you, when you said you read it in two afternoons I was so surprised but then when I started reading it, it’s so easy. You can read a whole chapter in under an hour very easily.
Sean Casey: 56:11
Yeah, and I was, I don’t know, I suppose the setting where I was at the time, the book, everything, it’s just one of those moments that I look back on, like that penny drop I said earlier on. It was so enjoyable. And a great read and I followed Cole as well on LinkedIn and seen some of the stuff that she’s [inaudible 00:56:33], some of her talks and presentations and it’s great. It’s great to keep learning [inaudible 00:56:39].
Kirill Eremenko: 56:39
Fantastic, all right. Well, Sean, thank you so much for coming on the show today. It’s been a pleasure.
Sean Casey: 56:45
Nice one man, thank you very much for having me and thank you to all the community, you’re great.
Kirill Eremenko: 56:55
So there you have it, thank you so much for spending this hour with us. I hope you enjoyed the conversation with Sean and got lots of valuable take aways. I actually had read his story, he sent it to me in the email before the podcast, so I knew lots of, many parts of it, but at the same time, during podcast, I found myself listening and mesmerized by how he was describing the things that led him to be where he is now.
Kirill Eremenko: 57:21
Every story is unique, every story is so interesting and thank you very much, Sean, for coming on the show and sharing your story. My favorite part probably was the advice that Sean gave at the end. Start small. It’s such valuable advice. Data science is such a broad field. Doesn’t mean if you’re into data science you have to do machine learning, computer vision or artificial intelligence. Don’t have to be an expert Python programmer, you can go into data visualization, or you can go into machine learning and Python. Or you can go into data preparation and SQL and databases. Or you can go into data science leadership and management and things like that.
Kirill Eremenko: 58:01
There’s lots of areas to get into data science, and by starting small, you reduce the downside. Basically, you don’t invest three years of your life into a degree that might not be exactly that part of data science that you want to be doing. So, starting small, trying out a few courses, understanding what you actually like about this field is a great, great thing. And of course, talking about the data science community, that was fantastic. I love everybody in the data science community. It is so friendly.
Kirill Eremenko: 58:31
As usual, you can get the show notes for this episode at SuperDataScience.com/383, that’s SuperDataScience.com/383 where you will find transcript for this episode and any materials we mention on the podcast.
Kirill Eremenko: 58:44
And if you found this episode inspiring, educational, motivational, that it challenged you, that it approached you to think in a different way, then share it with somebody you know. Somebody who might need that extra boost of motivation or inspiration to keep going with their data science journey. Very easy to share, just send them the link, SuperDataScience.com/383.
Kirill Eremenko: 59:04
And on that note, I look forward to seeing you back here next time. Until then, happy analyzing.
