Kirill Eremenko: This is episode number 323 with Top Rated Freelance Data Scientist Wesley Engers.
Kirill Eremenko: Welcome to the SuperDataScience podcast. My name is Kirill Eremenko, Data Science Coach and Lifestyle Entrepreneur. Each week we bring you 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: This episode is brought to you by SuperDataScience. Our online membership platform for learning data science at any level. We’ve got over two and a half thousand video tutorials, over 200 hours of content and 30 plus courses with new courses being added on average once per month.
Kirill Eremenko: And you can get access to all of this today just by becoming a SuperDataScience member. There is no strings attached. You just need to go to www.superdatascience.com and sign up there, cancel at any time. In addition, with your membership, you get access to any new courses that we release plus all the bonuses associated with them. And of course there are many additional features that are in place or are being put in place as we speak, such as a Slack channel for members, where you can already today connect with other data scientists all over the world or in your location and discuss different topics such as artificial intelligence, machine learning, data science, visualization and more, or just hang out in the pizza room and have random chats with fellow data scientists.
Kirill Eremenko: Also, another feature on the SuperDataScience platform is the Office Hours where every week we invite valuable guests in the space of data science and interrogate them about their techniques, about their methodologies in the space of data science. And you actually get a presentation from the guests and you get an opportunity to ask Q&A at the end. And in some of our office hours we just present some of the most valuable techniques that our hosts think are going to be valuable to you. 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.www.superdatascience.com secure your membership and take your data science skills to the next level.
Kirill Eremenko: Welcome back to the SuperDataScience podcast ladies and gentlemen. Super pumped to have you back on the show because today I’ve got a very special guest. A very special surprise for you. Wesley Engers is a top rated data scientist on the world’s biggest marketplace for freelancing, which is Upwork. So if you haven’t heard of Upwork before, it is indeed the largest online marketplace for freelancers. Just some stats. So it has approximately three million jobs are posted on Upwork annually, and their total value is about $1 billion. One billion with a B. And that is a huge amount. As of 2017, Upwork had 14 million users in 180 countries. And as you can imagine, that has most likely grown very substantially since 2017.
Kirill Eremenko: So Wesley is one of the top rated data scientists on Upwork, and he was actually featured on Upwork homepage a few weeks ago. Maybe you can still find him there. And that’s how I found out about him. I went on Upwork, that’s where we hire quite a few freelancers for SuperDataScience, and I saw his profile there and I saw that he’s a data scientist and I knew that moment I had to bring him on the podcast to share with you, our amazing audience, the benefits, the pros and cons, the tips and hacks about freelancing and data science. So if you’ve ever wondered about it now, you will be able to make an informed decision whether a career in freelance data science is the right career for you or not.
Kirill Eremenko: So here are some of the topics that you will learn about on this podcast. First of all, we talked about freelancing as a data scientist and how Wesley got into this space in the first place. Then we talked about how you can get started if you want to dive into freelance data sciencing yourself. We talked about things like when to increase your rate, good clients versus bad clients, how to delight your clients and keep them happy. How to think about their needs. We talked about the tools that Wesley uses. He’s an avid R user. So you’ll learn a bit of things in that space here as well. We talked about some sample projects, and very exciting sample projects from a diverse range of fields ranging from medical project in the medical space to projects in oil and gas industry and so on.
Kirill Eremenko: We also talked about freelancing versus full-time work, and how to choose a career in one or the other or how to combine both. And at the very end, so if you’re really serious about getting into freelancing, especially about Upwork, make sure to listen to the end because we gave two tips, one tip each about getting into the space and those are make it or break it tips. You don’t want to miss out on those.
Kirill Eremenko: So those are just some of the topics that we touched on. This is the podcast where you can learn a lot about freelance data science work. So if that’s the space you were interested in or you’ve ever considered, then off we go. Without further ado, I bring to you experts and top rated freelance data scientist Wesley Engers.
Kirill Eremenko: Welcome back to the SuperDataScience podcast ladies and gentlemen, super excited to have you on the show. Today, we’ve got a very special guest, Wesley Engers calling in from San Jose, California. Wesley, how are you going today?
Wesley Engers: Good. How are you doing?
Kirill Eremenko: Amazing. Doing amazing. I’m in Brisbane, Australia, and I was actually, as I mentioned to you, talking to someone else from California, from LA, and it looks like you guys are having some cold weather and rains these past couple days.
Wesley Engers: Yeah. I think it’s really getting into the winter season here for us, that means rain.
Kirill Eremenko: No, at least it’s not snow, right?
Wesley Engers: Yeah, exactly. You have to go up to the mountains or Tahoe for that.
Kirill Eremenko: Yeah. I think if it snowed in California, there’d be so many traffic jams.
Wesley Engers: When we get this little amount of rain, just the traffic around here in Silicon Valley is already pretty bad, but just get a little bit of water and then nobody knows how to drive anymore.
Kirill Eremenko: Yeah. In winter countries, I know that they change tires and people would be totally not prepared for changing tires there. Crazy. Have you always been from San Jose? Have you lived there all your life?
Wesley Engers: No. I’ve been here since college. Before that, I actually grew up in the Ann Arbor, Michigan, in the Midwest of the United States. Very, very nice out there as well, weather is not quite as good though.
Kirill Eremenko: Okay. Is that why you moved, for the weather?
Wesley Engers: Yeah, for the weather and for school. I really liked the culture and the vibe here in Silicon Valley. It’s always really interesting being on the cutting edge of technology and all the new internet and tech stuff that’s coming out. So I think that was another reason to move here beside the weather.
Kirill Eremenko: Exactly. Exactly. Like proximity, right? Proximity is power. You want to get something, you want to learn from someone, you got to be close, like physically close.
Wesley Engers: Yeah. Try to be close to the best. I think as far as technology and new innovation, Silicon Valley is the best in that.
Kirill Eremenko: Yeah. So has it been worth it? Have you found any useful mentors or people that you aspire to be like or learn from?
Wesley Engers: Yeah, I’ve definitely connected with a good number of people out here. First off, through the college I went to, Santa Clara University. Math department there was great for me, met some great people there, great professors. And then going into the workforce and as I’ve gone through the consulting that I do, definitely met some really great people, really intelligent people that you can learn from.
Kirill Eremenko: That’s very cool.
Wesley Engers: Yeah.
Kirill Eremenko: Awesome. I love that when that pays off. It might be a difficult move to make and even consider moving to across the country, but eventually it all works out. Speaking of technology, cutting edge, Wesley, you are very interesting because one day I opened Upwork… So for those who don’t know, Upwork is the largest online marketplace for freelance workers. So there’s supply and demand side, businesses go there to hire freelancers. Freelancers go there to provide services, largest in the world by far. And with the SuperDataScience, we do a lot of work through Upwork. We’ll hire a lot of people. One day I open Upwork page and I see your face on the front page. Like you’re standing there, this green background, “Wesley professional data scientist”. I was like, “Whoa, Upwork is featuring a data scientist.” Congrats on that, that’s a huge one.
Wesley Engers: Yeah. Thank you. It was a lot of fun. A little bit surprised to get that. I knew I was a top-rated and doing good work on there. But it was really nice to get a call from, I guess, their marketing department.
Kirill Eremenko: Tell us how that happened. So you’re just sitting there in your office one day or like coworking space, where was it? You get this call, you’re like, “Oh, what? You want to feature me?”
Wesley Engers: Yeah.
Kirill Eremenko: How did it happen?
Wesley Engers: It was actually an email. It wasn’t a call actually. They just sent me an email. I think maybe kind of similar how you reached out to me and sent me an email asking, “Oh, we’re updating our homepage and looking to feature freelancers in each of our major categories. And we like you to be the data science one. And here’s sort of the stuff that that’s going to entail. Would you be interested? And if so, let’s set up a call, make sure everything’s in alignment.”
Kirill Eremenko: Go for photos, go to the studio.
Wesley Engers: Yeah. Yeah. Did the phone call, a little video conference, and I guess they liked me well enough, so then they had a photo shoot day, and that was up in San Francisco. So pretty convenient for me. About an hour, hour and a half from my place. Drove up there and yeah, I did the photo shoot, full thing. They get the wardrobe people and the makeup artist and [crosstalk 00:11:32]
Kirill Eremenko: Oh, nice. I bet it felt like a celebrity. Not often do data scientists get a makeup artist.
Wesley Engers: Yeah. And then they asked for some more information so they can do those features on you. I don’t know if you just saw the… So I’m on the homepage and then they’re also going through and doing these campaigns to feature various freelancers where they do [crosstalk 00:11:59]
Kirill Eremenko: Oh, no. I didn’t see that. That’s exciting. Wow, you got featured in those as well. They’re very cool. Very cool.
Wesley Engers: Yes.
Kirill Eremenko: I realized you’d be so excited about this opportunity, but to me even hearing that sounds super exciting because basically what that means is that Upwork recognizes as data science is one of their major categories. Did you learn anything more about that? How did you feel about that part of the news?
Wesley Engers: Well, I think there’s been a lot of hype around data science and I think there is a lot of advance in AI and machine learning, and data science is a relatively new field. So there’s not tons and tons of experience in it. So I think it makes sense for them to highlight such a growing market. But yeah, it definitely feels good to be in a growing category where there’s good potential for longterm growth in the future because it means there’s going to be good demand for my skills, I think.
Kirill Eremenko: Fantastic. Okay.
Wesley Engers: Data science in general skills. Yeah.
Kirill Eremenko: So let’s dive into that a bit. How long have you been doing data science for?
Wesley Engers: I’d say probably around since I graduated college around 2011. Definitely going through grad school and stuff, I was always doing this consulting stuff on the side as I was doing my graduate studies.
Kirill Eremenko: So about eight years or so?
Wesley Engers: Yeah [crosstalk 00:13:40] eight plus years.
Kirill Eremenko: How about Upwork? How long have you been on Upwork?
Wesley Engers: Well, I’ve been on Upwork since they started. I actually originally joined through a platform called Elance.
Kirill Eremenko: And then they got acquired by Upwork [inaudible 00:13:58]
Wesley Engers: Actually. So there is Elance and there was oDesk, and they were both freelancer marketplaces. I think Elance, to me, was a little more towards higher quality stuff, and o-Desk was more like admin, virtual assistant type stuff. But they did a merger. And the merger between the two companies, and then they rebranded and came out as a Upwork I think.
Kirill Eremenko: Oh, that’s how it was.
Wesley Engers: [crosstalk 00:14:23] so after the merger.
Kirill Eremenko: Got you. Okay. Got you. But how long have you been active on Upwork? All this time or is it just [inaudible 00:14:33]?
Wesley Engers: Yeah. No, all this time since they launched. Don’t hold me to these exact dates, but I think it was around 2015, 2016 timeframe that they came out, and I’ve been active on the site ever since. It has definitely ramped up in terms of exposure within maybe the last year and a half or so, getting more clients. And I think that’s probably just partly in nature of being in the business longer and just takes that time to build up that reputation and have clients recognize you.
Kirill Eremenko: Okay. Got you. Well congrats on the huge success. So for our listeners, here are some mind blowing stats. As a part-time freelancer, something that you do on the side as I imagine while you’re studying, while you’re now running your own business, mentor analytics and other things, you’ve managed to earn $60,000, 60,000 US dollars on Upwork by completing 115 jobs, working just over 469 hours. That’s the per hour jobs. There’s probably is more hours, of course, in the fixed payment jobs. And your current hourly rate, so if somebody on Upwork wants to engage you as a data scientist, right now your currently hourly rate is $180 per hour. Mate, congrats on that. Those are like some really cool, impressive stats. Very inspiring.
Wesley Engers: Yeah. When you read them that way, they sound really, really great. Always good to hear.
Kirill Eremenko: Yeah. And so tell us about your experience working on Upwork. No, you know what, let’s start earlier. How did you get into data science in the first place? Because it sounds like you already in college knew that, “Oh, this is something I want to do”. You were already doing some part-time work in data science back since 2011. What did you study and how did you come in to being in the data science space?
Wesley Engers: Yeah. I studied math and finance as a double major in math and finance at Santa Clara. And then after that I went to Claremont Graduate School and got a master’s in mathematics. So I definitely come from the math side of things rather than… I think computer science is another pretty common background to get into data science. I mean there’s some other backgrounds too, but I think those are two of the most common.
Wesley Engers: So I’m definitely more from the math stats side of things. I’ve also always been interested in business and finance, that’s why I studied finance in undergrad. So this is sort have been combining those two. That’s sort of why I started doing this part-time is I’ve always believed in sort of multiple income streams. So in addition to doing a main job or something, these days gig economy is sort of really big. But I did start to get into that, I think, maybe before it really took off, doing some of these side consulting projects and also do a lot of math tutoring throughout undergrad and grad school at the same time. So to me, part of it has always been about building these multiple streams of income.
Kirill Eremenko: What was your first exposure to data science? Because mathematics and your statistics is part of data science, but it’s also a separate thing. You could have just become a statistician. When was that first time when you decided, okay, this is what I’m really passionate about, this is what data can do for business? Do you remember that first exposure, that first project that you did?
Wesley Engers: I think data science is such a broad field that it’s, to me, very much a broad spectrum. So it’s very hard me to say like, “Oh, this thing was definitely in a data science category and this other thing, oh, that was just in the math category.” I definitely do remember, I guess probably… I think it was my first job on Elance actually. The guy ended up being a really good client, and I’ve worked with him over several years, still occasionally keeping contact and send me some work every now and then. He wanted me to look at some data and just build some graphs for him.
Wesley Engers: I mean, this was in pretty much Excel. There was definitely some data checking components to it, make sure that he’s transferring stuff over, I think, from some PDF files. And that’s pretty much just all manual work. Just verifying that the numbers in the spreadsheet are right and then building out these various graphics for him to use. And by graphics, I mean like line charts and bar charts and that sort of thing. So nothing too sophisticated here.
Wesley Engers: But I think just getting that first job and realizing that somebody is going to pay you money to do this directly, and that this is bringing value to them, or at least more value than they’re paying you, I think that was really inspiring for me to be like, “Oh, people actually want to pay to use my skills.” And I really enjoy working with the numbers and trying to convert all of these numbers, which a lot of people aren’t really quantitative. They don’t want to just read a and have understanding from that because that’s really hard. They want these nice visuals, these graphics so that they can look at something in five, 10 seconds and have a much better understanding rather than pouring through data for an hour or so to get the same kind of insights.
Kirill Eremenko: Very cool. Very cool. So kind of your first entry to data science was through business intelligence, and then slowly developed. When did you start getting into more modeling type of data science? More like the machine learning side of things?
Wesley Engers: Yeah. I will say a lot of the work that I do is, I think as you said, maybe more towards the business intelligence side of things, but modeling is definitely important thing.
Kirill Eremenko: What would you say is the percentage ratio? Because I’m looking through your past jobs on Upwork. By the way, everybody, we’ll share a Wesley’s profile on the show notes, and you should go and check it out because the jobs actually have the titles, some of them you can go in and see the descriptions, what people wanted from the job. So the top jobs, the most recent ones are predictive model and algorithm for sports analytics, predictive model for natural resource supply and demand modeling, statistician for analysis, statistical advice about estimating cost curves, ad hoc stats and data consulting. It looks like most recently it’s being modeling. So what would you say is the ratio between business intelligence jobs and predictive modeling or machine learning jobs?
Wesley Engers: Yeah. I don’t really keep track of the percentages.
Kirill Eremenko: Very strange for a data scientist not to keep track of percentages. But what’s your gut feel? If you’re looking back, would you say it’s 50/50, 70/30?
Wesley Engers: Yeah, I think 50/50 sounds about right. But the thing is like the modeling jobs, I guess maybe here’s where it kind of gets skewed. Is in order to do the modeling, you have to do all the pre-cleaning work beforehand. So, regardless of what you’re doing, whether you’re doing a modeling job or you’re just building graphics for somebody, you still have to go through and clean the data. And I think probably as your viewers have heard a lot before, right, 70, 80% of a data scientist job is not building models, it’s cleaning data and getting it organized so that you have no missing values, that you don’t have duplicate values, so that you have the right features that you want to put into your model.
Wesley Engers: All right, that’s like 80% of the job. And then even if I’m doing a modeling job, then the last 20% or whatever, 10 to 20% is actually going to be building that model.
Kirill Eremenko: Got you. Makes sense.
Wesley Engers: Yeah. But I’d say like 50/50 because a lot of companies, I think, want actually more algorithms or various metrics. And metrics generally ended up being some kind of weighted average that’s really informed by industry expertise as to which factors are important. Now you can do some data verification here and there to make sure that, oh, if they say, “Oh, this factor is important.” Right? You want to go through the data and have some support and verification of this fact.
Wesley Engers: But I think generally if people have been in a particular industry, they’re very knowledgeable about that industry, and they actually do have a pretty good idea as to the key factors that are going to influence their decisions.
Kirill Eremenko: How about you? You’re doing these jobs across sports analytics, mining, oil, medicine, all these different industries. Bio statistician, I can see a job here. Crazy variety, very exciting. Is it hard for you to develop that domain or industry expertise?
Wesley Engers: Yeah. I think that’s always the trade off. So yeah, I’m very knowledgeable, I would say, about data how data variables relate to each other and what kind of models are appropriate based on the kind of data that you have. And what kind of visuals are appropriate based on the data types that you have. And I think getting a basic understanding of domain expertise, I really enjoy doing it, and I can get a lot of that actually from my clients. So the clients, when they come to me, generally have a lot of domain expertise already. That’s the reason they’re in their particular industry, but what they don’t have is they don’t have a lot of knowledge about how data works or how to build statistical machine learning models. So that’s what they come to me for, right?
Wesley Engers: And through our initial calls about what are their objectives for the project, what do they want to get as an output from me, what kind of data do they have available? There is a fair bit of back and forth communication between us to, I think, at least keep me in the loop about where’s the domain expertise coming in and they can give me a good information on that, and then I can do a pretty good job of building that into any model that I create.
Kirill Eremenko: Okay. Got you. Very interesting. So tell us a bit about the process. You said a client contacts you, then you have some calls, walk us through the steps you take to onboard a client and understand what are their actual needs.
Wesley Engers: Yeah. I think you the onboarding and communication step is very, very important for whatever kind of project that you’re running, whether it’s in a big company or like I do on the freelance side of things. You definitely want to figure out what is their objective for the project? What kind of output are they expecting? What kind of timeline are they on? What kind of data do they have available? And that one’s probably going to take a little longer. First off, once you establish those objectives, then you’re going to need to get access to the data somehow. Maybe that’s them sending me the data in a CSV format. Maybe it’s connecting to their database. Clients I work with tend not to have massive, massive amounts of data, so they can usually manage a CSV file. It seems to be the preferred method.
Wesley Engers: Once I take a quick look at that data, I’ll probably have some clarifying questions about, “Oh, how was this data gathered?” Just particular specific data questions. And then I’ll probably outline the steps that I recommend going forward. Sometimes there will be two or three, maybe different options for them to choose between based on how much time they initially want me to spend on it versus how deep they want me to go. The deeper that I go and the more sophisticated model I build, the more time I’m going to spend with it and the more expensive it is going to be for them.
Wesley Engers: They need to decide where the balance is there for their particular needs. Once they let me know about that, then that’s when you start really get into the data cleaning and the model building. And from there, you sort of go to the deliverable, which is either a presentation or a model or a dashboard, depending on their particular needs.
Kirill Eremenko: Okay. Okay. Got you. I see that you have both fixed jobs and per hour. And for those who are not familiar with Upwork, basically it’s two types of jobs. The client might say, “Okay, I have this budget, this is how much I’m going to pay.” Or you might agree on a price with the client together in advance like $200 or $1,500 or $5,000 for this job, or it can be per hour. So what would you say is the most common type of work you’re hired for? Is it per hour or is it fixed price work?
Wesley Engers: I’ve mostly done per hour work. I think I am trending towards doing more fixed price jobs. But I think there’s room for both of them. The advantage of having an hourly rate is the client can be a little bit more flexible and respond based on the feedback that I tell them. After I’ve done some data exploration and got started on the project, I can tell them, “Oh I found this, this and this and those are really interesting. Do you want me to dig into these facts a little bit more?” And on an hourly job, they’ll have more flexibility from me in doing that. But the advantage of the fixed price thing is, I think on the cost side, they don’t have any variable costs anymore, right? It’s a fixed price, they know exactly how much they’re going to pay, they know exactly what they’re going to get.
Wesley Engers: So when I’m doing a fixed price job, I make extra sure to have really clear end deliverables so that both the client and I are on the same page, and we’ll both be happy with me on terms of how many hours it’s going to take me to complete this thing, and them on the very particular deliverable that they want. So if they know sort of what they want at the end of the day, then a fixed price job is a good way to go. But if they’d like some more flexibility to change up the project and make edits to the needs of the project as it progresses, then I think the hourly job can be better.
Kirill Eremenko: Tell us about good clients versus bad clients because as freelancers can be good and bad, I imagine clients can be good and bad. So how do you select or vet your clients seeing that, okay, this is a person or a business that I do want to work with, versus this a business that I don’t feel right working with where you can see that there’ll be problems down the line.
Wesley Engers: Yeah. I think I’ve been pretty fortunate I’ve not had too many what I would consider bad clients. Luck or just the vetting process hopefully helps with that. You want clients who communicate really well. They respond in a timely manner to any questions that you have on their project. You want clients that are going to be detailed. So when they’re posting up their job they should have a pretty detailed post about exactly what they need and what they want. If you just post up, “Oh, I have some data, I want some analysis on it.”
Kirill Eremenko: That’s the worst kind of job.
Wesley Engers: I’m like, well, that could be like anything. I don’t know what you want there. But if you’re like, “Oh, you know what, I’m an online retailer, I have a bunch of customer data and I want to figure out who are my top customers and what kind of characteristics do they have.” Then that’s a pretty good something to build off of, right? It’s pretty clear that they want more information on their customers. They have some data on it.
Wesley Engers: So detail oriented, good communication, and I think… Well, also they want to value your expertise, right? They’re hiring you because you’re knowledgeable in your field. So I don’t want clients who are going to skimp. I don’t want people who are looking for the cheapest freelancer. That’s why my hourly rate is where it is. It’s because the only kinds of people I want contacting me are people who value my expertise, and hence they’re going to show that value by saying, “Oh yeah, I think it’s worth it to pay this guy $180 an hour.”
Kirill Eremenko: Got you. So like your hourly rate acts as a filter, you’re already like just only letting through high quality clients and minimizing the amount of additional filtering that you’ll have to do by looking at the job description and things like that.
Wesley Engers: Yeah. I think my price is definitely a filter in there to help get those high quality clients. I’ve done it in the past and people have asked for a lower hourly rate or a discount. And I feel pretty much all of my bad clients have come from the people that I have agreed to give discounts to. Not everybody in the past that I’ve given discounts to has turned out to be a bad client. But I don’t think I’ve really had good client or clients who haven’t negotiated some kind of discount or put up some of these sort of red flags here. Yeah. They’re always the ones that turn out.
Kirill Eremenko: So it’s basically best to stick by your decision and use that price as a filter?
Wesley Engers: Yeah.
Kirill Eremenko: Got you.
Wesley Engers: Yeah, I’ve definitely gotten a little burned from not doing that and not following my own advice.
Kirill Eremenko: And speaking of your hourly rate, it’s grown over the years, right? So back in 2015 and so you started like at $35 and went up to 40, 50 120, 150. 180. When did you feel confident? Like for somebody who’s a freelancer now, who’s looking to get into freelancing, when is the right time? What’s a good starting rate, first of all, on Upwork, for a data scientist in your opinion? Somebody who’s brand new to Upwork. And second, when is the right time to increase your rate?
Wesley Engers: Yeah. I know, I think it’s always really hard trying to pick a starting rate. So if you’re just starting out, yeah, you’re definitely going to be on the lower end because you need to build up that reputation. I think something in the probably 30 to $50 range per hour is probably a reasonable starting rate for the data science category.
Wesley Engers: How do I decide when to raise my rate? It generally has to do with the amount of demand that I have for my work, and when I feel that I’m getting too much work to do at a given hourly rate, then I raise my rate and like, oh, well, I’m getting too much work right now, I have too many clients. Okay, so that means it’s time for me to raise my rate, get more hourly or get fewer clients for a higher hourly rate.
Kirill Eremenko: Okay. Okay. Got you. That makes total sense. So in a similar fashion, if you don’t get enough work, it’s always okay to drop your rate a bit so that you get more work and then you’re back on that trend.
Wesley Engers: Yeah. Yeah. If you’re looking for more work at a given time, yeah, drop your rate a little bit. I think a lot of it does depend on your personal financial situation, as to how much you want to do that. Sometimes I’m like, “Well, I don’t have a lot of clients right now, but I don’t really care. I want more free time anyway.” So works for me.
Kirill Eremenko: Yeah. Got you. Okay. So tell us a bit about the tools. What tools do you use for your data science projects?
Wesley Engers: Yeah, I’d say the main tool that I use is R.
Kirill Eremenko: Hey, R fans out there.
Wesley Engers: Yeah, R fans. I have the mass stats background, so I went into R. I think I definitely, I first used that I think in a class or two in my undergrad and maybe a class or two in grad school also required me to use R. But I actually didn’t really get into it until I took… I did the Johns Hopkins 10 course series on Coursera.
Kirill Eremenko: Oh yeah. I did one of them. I did the R one. R programming one on that one. Yeah, it was really good.
Wesley Engers: Yeah. So that entire 10 course series is in R. So I’d say that’s really where I got my start in using R on a regular basis. And then once I completed that, I started doing consulting stuff in R because they found it to be a very good tool.
Kirill Eremenko: What else?
Wesley Engers: So R I would say is my primary-
Kirill Eremenko: Data cleaning.
Wesley Engers: … data science language-
Kirill Eremenko: Data modeling, all that stuff.
Wesley Engers: Yeah. It’s so flexible. I also love that it’s free and it’s open source. You can find whatever packages that you want. There’s a good community around it. So anytime I get stuck on anything I can just go to Google and probably come up with stack exchange or stack overflow or something like that. And you can solve your problem pretty quickly that way.
Kirill Eremenko: Yeah.
Wesley Engers: And I also end up using Excel a good bit. A lot of my clients are not as data savvy, so sometimes if… Especially if they have a tool that they want built rather than just a report at the end, a lot of them will want it in Excel so that their people, once the tool is built, can go ahead and maintain the tool and update the tool with data. I also build things in Excel for them. Sometimes that does actually mean I will go build say a regression model or something in R that’s a little more sophisticated than Excel could build. And then once I have that model, I’ll sort of basically just import that model into Excel. And at that point, that model is fixed, which is a little sad for me because I’m like, “I wish it could be a little more dynamic.” But they needed in Excel, so that’s kind of what I have to do.
Kirill Eremenko: What do you mean you import it to Excel? Like you re-code it in Excel or you take the actual R results and just put them in Excel? What do you mean by that?
Wesley Engers: Yeah, no, you just take the R results and you put them in Excel.
Kirill Eremenko: Okay. Got you. Okay.
Wesley Engers: Yeah like, a very trivial example, but if our equation ended up being Y equals three X plus five, or something, just kind of hard code three X plus five in to-
Kirill Eremenko: Oh, the coefficients. So you move the-
Wesley Engers: Yeah, I just move the coefficients. I mean Excel can build those kinds of basic linear regressions, but for a more sophisticated model-
Kirill Eremenko: Got you. So basically you wouldn’t be able to retrain the model in Excel, you would have to retrain it in R again?
Wesley Engers: Yeah, yeah. I’d have to retrain in R, if they wanted an update on the-
Kirill Eremenko: That’s very cool. I love that you mentioned that. It takes a lot of humility to mention. To do work in Excel or to be confident to do work in Excel, why is because there are data scientists out there that will swear on the lives that Excel is the worst tool for data science. They’ll probably stone you to death if you mention something like that in front of them. So it’s really cool. I think it’s a good point that you got to adapt to the client’s needs, right? Like R is great, but what does the client need?
Wesley Engers: Yeah. No. I think that’s absolutely been my approach and my philosophy is, yeah, you know what, I love to use R all day long, but the reality of it is what does the client need, what does the client want? And that’s what I need to deliver on. And I think it is really easy. When you get into data science, you’re like, “Oh, I want to build the most awesome, sophisticated, complex model that I can. And it’s going to have this great accuracy rate and predictive power and all of that.” But sometimes it’s so complicated to understand, the client like-
Kirill Eremenko: They’ll just throw in the trash. They’ll just like bin it.
Wesley Engers: Yeah, they’re just going to throw it in the trash. Actually I think this reminded me of… Was it back in 2009 or something, Netflix had the Netflix plot prize for proving their recommendation algorithm.
Kirill Eremenko: Very famous thing. On Kaggle.
Wesley Engers: Yeah.
Kirill Eremenko: No, or maybe not on Kaggle, but yeah, definitely heard of that one. Yeah.
Wesley Engers: Yeah. They had this Netflix prize. I don’t know how much it was. A lot of money I’m sure.
Kirill Eremenko: I think it was a million bucks. I think it was actually a million bucks.
Wesley Engers: It was a million? Okay. That’s what I was thinking.
Kirill Eremenko: Yeah.
Wesley Engers: And yeah, they didn’t end up using the winning algorithm because the winning algorithm, while it was better, it wasn’t that much better and it was just so much more resource intensive that it wasn’t practical to use it. When you’re building these models, you have to balance a lot of different resources.
Kirill Eremenko: So basically the winner didn’t take into account that Netflix needs to do this all the time, real time. It has to be continuously running on their servers. They’re going to scale to a massive company with… Right now it’s the biggest subscription company in the world, with the biggest membership, full-stop. Obviously that gives certain constraints to the types of models that would be useful to them.
Wesley Engers: Right. I think the contestants, they won and they absolutely follow all the contestant guidelines. So I think part of it also falls on Netflix for not realizing that they needed to put these additional constraints for their problem.
Kirill Eremenko: Yeah. Yeah. That’s right.
Wesley Engers: Yeah.
Kirill Eremenko: Okay. Okay. Very interesting. Okay. So R, Excel, very cool. What do you use for BI?
Wesley Engers: Well, I can say I actually don’t do… Those are the main tool tools that I-
Kirill Eremenko: Okay. So even for BI, you would use R and Excel? Like you’d create charts and visualizations because maybe you use Tableau or Power BI or something else. But it’s R and Excel mostly?
Wesley Engers: Yeah. R and Excel. Actually when I was working at Ed Symantec, I did some SQL stuff, but that’s not exactly… That’s still kind of the data pulling side of things. Yeah. So building visuals, I’ll definitely use ggplot2 and-
Kirill Eremenko: ggplot2 is good. Really good. Very good.
Wesley Engers: There’s a lot of other add on stuff with ggplot. Recently I’ve been doing some of the geospatial stuff with the oil dataset I’ve been working on. And I’ve used, I think it’s ggmap, so it can pull in Google maps and integrated with ggplot and it works.
Kirill Eremenko: Oh, that’s sweet.
Wesley Engers: Fantastic.
Kirill Eremenko: That’s sweet. Speaking of oil, let’s switch gears a bit and talk about some projects that you’ve done. Get us excited. What are some of the coolest projects you’ve done as a freelancer?
Wesley Engers: Yeah. I think you mentioned that oil and gas data, it’s still an ongoing project.
Kirill Eremenko: So what’s that all about? As much as you can share. Of course, I understand there’s certain caveat, so you can’t disclose everything, but whatever is public or wherever you can disclose.
Wesley Engers: Yeah, I try to keep it general here. Yeah, so client got some oil and gas data from drilling wells in Texas actually. And yeah, this one’s been, I’d say, a bit more open-ended than my typical project. So a lot of data exploration here, but definitely working towards building models for trying to figure out where to drill wells, what kind of geology is going to be predictive of high oil content for high production value. And I think possibly also working on the cost side of things as well like how maybe certain geology are going to be more expensive to work through. So you’re balancing the revenue side of things and the cost side of things. So what kind of different engineering stuff are you going to have to do for harder geology, I guess?
Kirill Eremenko: I’ve heard numbers around that it takes about 20,000 to $40,000 just to do one test drill. Just to drop the shaft into the… Or like make the drill and drop the shaft in there, to check if this is a good place. That’s 20,000 to 40,000 bucks. And they do dozens, if not hundreds of them, to find where to continue drilling to explore. And it still is not a guarantee. So if you are able to provide value, imagine how much money you can save the company like that.
Wesley Engers: Yeah. Having more intelligence on the geology and how the geology relates to the overall oil production, I think, is really important for the oil companies. And if you can say develop some kind of software or something to really help them with this, I think there’d be a lot of demand for that, or hope there would be any way.
Kirill Eremenko: That’s a big one. That’s a really big one. Do you have another one? Different industries, something that might be very exciting?
Wesley Engers: Yeah. Let’s see what I can think of here. I did do some stuff… I guess this is healthcare, more healthcare-related. They were, I guess, optometry manufacturer. So just doing a customer segmentation analysis here, they run a pretty extensive survey. So went out and surveyed a bunch of their customer types-
Kirill Eremenko: Sorry, this wasn’t you going out? They went out, you just got the [crosstalk 00:48:33]-
Wesley Engers: No. They had already done the surveys, so they provided the survey and basically summarized results of the survey to me. They gave me the raw survey too, so I could see the questions and look for any [crosstalk 00:48:43].
Kirill Eremenko: Just to clarify for our listeners, all of your work you do either from your home or from a cafe or from coworking, do you ever go out to the client’s location?
Wesley Engers: I very rarely go out to the client’s location. I have made some client visits in the Silicon Valley area. I think that is one advantage for me being based in San Jose is… Yeah, I’ve had kind of some Silicon Valley type people, and I can go visit them because they’re going to be an hour away from me at most.
Kirill Eremenko: Got you.
Wesley Engers: I do work through Upwork and stuff as you mentioned. So like the photo shoot was very close for me. I’ve visited Upwork headquarters before, not exactly a client visit, but sort of in a similar nature. I think I’ve done one or two client visits in the local area.
Kirill Eremenko: Yeah. But I guess with the Mentor Analytics, the start-up that you have on data science consulting, that that would become more common? Clients that are hiring you, not through Upwork but through your own company would sometimes expect that you come on site and actually work with them directly?
Wesley Engers: Yeah. I would not be opposed to that. I think there are definitely benefits to having face-to-face in person interactions. You can definitely get more done.
Kirill Eremenko: Got you. Okay.
Wesley Engers: Yeah. It also comes down to a cost thing. From a cost perspective, it is much cheaper for me just to work here from a cafe or for my room or laptop, wherever that is, than go in person. I guess that being said, I did do several trips over the summer. Clients did fly me around to a couple of different places on the East coast.
Kirill Eremenko: Nice. Well, maybe that’s the thing was going to happen. As you grow, as you get more and more clients, and especially with your expertise, if you’re going to be saving tens of thousands of dollars with your analysis or creating additional revenue, it’s a no brainer for a client to fly you not just across the country, across the world, if they need to get you in front of the right people. If the job cannot be done remotely.
Wesley Engers: Yeah. If the job requires that I’m, and it’s saving plenty of money, then I think it’s definitely worth it and make sense.
Kirill Eremenko: All right. Got you. Let’s get back to this optometry example. Sounds really cool. So you had these optometry manufacturer that went around serving their clients, sends you a lot of stories and they’ll ask you to segment them, right? Like geodemographics segmentation.
Wesley Engers: Yeah. Do essentially or exactly a customer segmentation analysis. Basically most of the questions were kind of numeric or you could make them categorical and then make them dichotomous variables, or some kind of sliding scale, like scale type questions. And then from that, did PCA for standardizing and scaling the variables, and then went ahead and did a cluster analysis, K means cluster analysis, to go ahead and establish the appropriate clustering. I think I used the… What is? The elbow method for trying to pick out those particular optimal number of clusters.
Kirill Eremenko: What was the toll on those clusters?
Wesley Engers: Oh, it wasn’t. It was like five or six.
Kirill Eremenko: Five or six. Okay. Well and-
Wesley Engers: Customer profiles.
Kirill Eremenko: Were they surprised at the profiles? Because I love this when it happens, when you’re able to generate customer profiles, not by just talking to your customers and kind of having a gauge of what kind of people they are, what kind of businesses they are, but actually just through plain data. K means clustering, and these are five clusters, and this is what they are. How did the client react to that?
Wesley Engers: Overall, I think the client was pretty happy with it. Yeah, I think they were very happy and we went through, and you get the center of your clusters, right? And you figure out, what does that correspond to in terms of questions? Then you can kind of build up that demographic information, which I think is much more impactful to them. They’re like, “Oh, you know.” These kinds of people like, oh yeah, I know certain clients who kind of fit into that demographic, right? On that category, right? And then we have these other clients, they fit into that other one. Yeah. So I think it went over very well.
Kirill Eremenko: Nice. Very cool. Do you know anything about the results of that? How much money you saved them or what kind of initiatives they undertook from that? Or this was the end of your involvement in this project?
Wesley Engers: Yes, that was the end of my involvement with that particular project. I do sometimes wish that I would hear more after the end of my involvement about the particular impacts, but I think a lot of times what happens on the freelance consulting side of things is I’ll do my one particular part and I’ll make my deliverable, but then I don’t get to hear necessarily the impact of that down the line.
Kirill Eremenko: That’s the trade off of freelance work. If you’re working full-time in a company, you get to see the fruits of your labor and them making an impact. Consultants, freelancers, you do your part and you leave, and you often never find out what happened next.
Wesley Engers: Yeah. Now some of them, they’ll hire me on for some other project later. But yeah, a lot of them, yeah, you don’t very often hear the impacts of the work later on down the line.
Kirill Eremenko: Yeah. Got you. All right, well that’s been two, oil and gas, optometry. Give us one more. Let’s round it up to three. One more example from a different industry.
Wesley Engers: Yeah. Let’s see. Yeah. A while ago I actually did one for a company that was I guess going through some class action lawsuit stuff. I guess maybe this one I know maybe a little more impact. Yeah, I’m sure they were on the hook probably for at least several million dollars in terms of these employees trying to sue them. And I guess these employees were workers that would go out into the field and they would end up driving these trucks that are owned by the company and they had a GPS trackers on all of their equipment. All of their vehicles. So I was able to look through the GPS data to establish how long, at least at a high level, an average level, various employees were driving around. I think their whole lawsuit was really around that, oh, they weren’t being paid enough over time, and this sort of thing. So I did a lot of-
Kirill Eremenko: So basically-
Wesley Engers: … [crosstalk 00:56:32] and aggregation.
Kirill Eremenko: … the employees were in these cars and basically they were like, we’re not getting paid enough for all the overtime, and yet you have the data on their GPS locations where they moved around. Do you try to help the company prove that actually they were being paid enough?
Wesley Engers: Yeah, they were being paid enough. We have the logs of how long they were driving these cars and we can see where they went. I’m not going through every single one. They have general patterns and you look for the various cutoffs for indicators as you’re going through and cleaning the data, obviously. Yeah, I’m going to go through… You’re going to have some misclassification here and there, but when you’re going through thousands and hundreds of thousands of rows of GPS tracking data, this is what you have to do.
Wesley Engers: Yeah. So you can come up with these, basically how long were they driving on each day for a given employee ID and that sort of thing. I’m pretty sure what happened with that one was that they ended up settling out of court. But I think they settled much lower than the employees were probably hoping for. So I think that was definitely a win for the company.
Kirill Eremenko: Oh, very interesting. And a fair win, not just like the mistreated employees.
Wesley Engers: Yeah. No, I don’t think they were mistreating them at all. But yeah, I know. You can get into these labor disputes and [crosstalk 00:58:04]-
Kirill Eremenko: That’s very cool. It’s kind of forensics data science over there. Really nice. Nice project. And you got that one through Upwork?
Wesley Engers: That one was actually not through Upwork. It was through another consulting firm that I was working with at the time. A guy named Michael, his kind of a mentor figure as well. He was running his own consulting firm. And this was one of his clients and he’d often hire me on to do the data, the data science, data analytics, portion of any projects that he got.
Kirill Eremenko: How did you meet Michael?
Wesley Engers: Now actually there you go, see, Michael I actually met through Upwork.
Kirill Eremenko: There you go. Yes.
Wesley Engers: He did find me initially through Upwork, but he was also based in the Bay Area, so we were able to meet up and coordinate and everything.
Kirill Eremenko: Very cool. Very cool. So there you go. That’s what I was getting to that like as you build your profile, as you build your expertise, you get connections, you start networking through these projects, and eventually you come up like something really cool, like that comes through your way. That’s very, very exciting.
Kirill Eremenko: Mate, thanks so much for sharing. We have very little time after the podcast. I do have a few more questions to ask you. So main one, or actually two main questions. So your thoughts on freelance versus full-time work. So there’s people listening to this, they’re like, I’m actually full-time employee, I have my data science job, or I’m transitioning to data science. I want to get a full-time job in data science. There’s plenty of demand for that. Why would I ever go into freelance? Or they’re maybe on the fence in between like, oh, should I be freelancers or should I be a full-time data scientist? What are your thoughts? You’ve done both. Where does your heart lie?
Wesley Engers: Yeah. I think that’s a very personal choice. For me, I think no question, love being freelance. And the reason that I love being freelance is the flexibility and the amount of control that I have over the work that I do and the schedule that I can set. So this is probably more relevant maybe for somebody who has a family and they want to take care of their kids and they want to do work after the kids go to bed or do work while the kids are in school. But I think freelance can be really, really good for work life balance, if you know how to manage yourself correctly. You also have the danger of letting the work control you instead of controlling the work, in which case… I’ve definitely read about all these stories of people who are self employed and working on their own and they end up working 12 to 14 hour days because they’re trying to do everything.
Kirill Eremenko: You’re talking to one right now.
Wesley Engers: Yeah. Okay, there you go. But generally I love the flexibility to set my own thing, and I think I have the right temperament to do this. So it works really well for me. I don’t think there’s anything wrong with being a full-time employee for a small company, big company, startup, whatever suits you, whatever you enjoy doing.
Wesley Engers: Now, I think lot of benefits over there as well, you have a nice stable job, you know what you’re going to do, go to work five days a week, probably put in the nine to five. I guess depending if you’re big tech company or something it can vary, but maybe you end up putting in more, but you have a nice stable job to go to, you have a steady income, very predictable income, you have access to great healthcare. That’s definitely one of my challenges is [crosstalk 01:02:02]
Kirill Eremenko: The uncertainty.
Wesley Engers: … the healthcare. Yeah, well the uncertainty and healthcare is very expensive to get without being part of a big company. And yeah, I have to deal with self-employment taxes and all of these other things. Whereas if you’re doing data science at a big company, you just get to do data science stuff basically. And you’re probably going to be building more sophisticated models as a full-time employee, especially if you’re working at LinkedIn or Facebook or Apple or Amazon or Google, any of these big ones. They have the resources to get you to work on these much more sophisticated models, whereas I need to meet my clients’ demands.
Kirill Eremenko: Yeah, very true. And it’s always a strategic thing. You might decide to like… Always look at your career as a conquest, with a strategic eye, basically. You might decide, all right, I’m going to go work for Apple or Google or LinkedIn and learn there for two or three years, but then I’ll become a freelancer, or something along those lines. And speaking of which I wanted to ask you, what are your thoughts on having a full-time job for our listeners who have, or aiming for a full-time job, and doing freelancing on the side, like doing some data science work on Upwork or other freelance websites and just gaining that expertise or additional income from side funds. Do you think it’s a good idea or maybe that’s too much dilution of focus?
Wesley Engers: Oh, no, I think that’s a great idea. That’s absolutely what I did. I was doing freelance part-time as I was working through school, and then when I went full-time at Symantec, I was also doing freelance stuff in the evenings. I think it’s a great way to get your foot in the door, see if you like it. If you don’t like the freelance work, like you can’t deal with the client demanding that, “Oh, where’s my item, I need it by tomorrow.” Or something like that.
Wesley Engers: I think it’s a great way to test the waters. And I think it’s also a great way to get started on building your clientele, it’s a much less risky way to get started in freelancing. I think I would actually recommend most or all people try to get that full-time job and then work on building up the freelance side of things, if that’s the route that they eventually want to go.
Wesley Engers: You just have so much more stability, particularly at the beginning when you’re working in a full-time position where you can count on that paycheck day in and day out. If you’re just starting from scratch and you don’t have a job and you’re trying to do this freelance thing full-time, your income’s going to be all over the map, and at the beginning it’s actually going to probably be much more towards the low end of things, and then you’re going to start to worry about, oh, am I going to have enough for rent or food, healthcare, that sort of thing.
Kirill Eremenko: The stress.
Wesley Engers: Yeah, you don’t want that kind of stress at the beginning if you can avoid it, right?
Kirill Eremenko: That’s a good point.
Wesley Engers: I would definitely recommend [crosstalk 01:05:24].
Kirill Eremenko: And so for people who are transitioning into the space of data science, if for instance you’re a developer or you already have a good stable job somewhere, that could also be one of the strategies to transition into data science. Rather than jumping into a junior role as a data scientist, keep your high paying job as a developer or as whoever you are right now and see how you fare as a data scientist on Upwork or on the side. That could be another approach. You could, of course, get a job as a data science and do that, or you could while you’re still kind of looking for opportunities, full-time opportunity in data science, why not start up on the side and see how you go?
Wesley Engers: Yeah. Absolutely I agree. I think it’s all about… You don’t need to take one massive step to get into data science. I’m always about advocating the progressions, the little steps that it takes you to get anywhere. So yeah, do those side projects. Another potential thing is if you’re interested in getting into data science, you already have that full-time job as maybe a developer or something like that. You can talk to your boss and see if there’s maybe a smaller kind of data science project that you could work on for your current company. And that way they know that you’re interested in moving into data science. And also you get a little bit of experience working on a project you’re going to enjoy.
Kirill Eremenko: Fantastic. Fantastic. Thank you for that. And last thing, what I want to do is let’s give a piece of advice for those who are determined, decided, like I want to try out this freelancing thing. I’m going to go and create a profile. I’m going to get started with this. I’d love to hear one piece of advice from you for people, and I also give one, and I’ll start first.
Wesley Engers: Sure.
Kirill Eremenko: My piece of advice is with the times when you joined Upwork, and when our Upwork was starting out are long gone. Upwork is now the biggest online marketplace in the world. And people listening to this, I’m really excited that whoever’s listening you’ve listened to this part of the episode because this advice might change the course of how you’re going to go about this. Upwork gets, and this is like recent research. I have friends who are excited, who want to be online freelancers, not in the space of data science, but I’ve basically worked with a few people to help them get started in Upwork. And I’ve done this research quite thoroughly, and Upwork gets 10,000 applications, wait for it, per day. 10,000 applications per day.
Kirill Eremenko: So there is no way in the world that they can accept everybody because otherwise the marketplace would be overblown. And so they have a very rigorous process of qualifying people. So first it’s an artificial intelligence that goes through your application, then it’s a person that goes through your application. So my advice is persistence. If you really have decided for yourself you want to be a freelance data scientist on Upwork, apply, and I guarantee you that your first application will be denied. Second one probably as well.
Kirill Eremenko: They will deny you up to three or five times just to see if you are really serious about it. Then I think that’s how the algorithm works. It just denies everybody right away, then okay if you could persist. And so you can apply up to like… There’s unlimited number of times you can apply, but people apply eight or 10 times before they get approved. I recommend, we’ll link to a few blogs in the show notes where you can read more about this on how to get your profile approved on Upwork, what people have gone through, some advice and so on. But basically persistent is key. It will take you a few weeks, even a few months to get approved. You have to keep changing your application, tweaking it, adjusting it to make it more user friendly or make it more appealing to them so that they can see that you will add value to marketplace, but you need to keep pushing.
Kirill Eremenko: If you give up after the first time, you think that, oh, well they denied me, then you’re not going to get on there. So make sure to be persistent. That’s my advice.
Wesley Engers: Yeah. I think you’re absolutely right. It has been many years since I first started on the platform. So in that sense I was lucky in getting in early. I think something that would be really helpful is trying to have a unique offering. I don’t think I’m your typical data science person, right? I tend to have a unique offering in terms of I think much more of an emphasis on the mass stats background and really making high quality deliverable to the client, right?
Wesley Engers: I’m not all about sophisticated model. I’m about the simple model that meets the client’s needs. So I think having a unique offering, and as much as possible having a unique skillset, is going to help you get into whatever freelance site you’re doing, whatever site you’re on, it’s going to help you stand out, get into the site and also get your first client.
Kirill Eremenko: Fantastic. Love it. Great advice. We are really short on time, actually run out of time. So Wesley, I want to say a huge thank you for coming on the show, hearing your expertise and knowledge. And before I let you go, can you please share with us, with our listeners, where they can get in touch, connect with you, maybe somebody is looking for some advice on their freelancing career and would love to have you as the mentor or maybe there’s some companies that are listening and really urgently need a data scientist and would love to fly you to, I don’t know, Jamaica or somewhere.
Wesley Engers: Yeah. You can definitely check out my website www.mentor-analytics.com. There’s a dash between the mentor and analytics. So mentor-analytics.com. You can find me on my Upwork profile as well, just Google my name there, or you can find me on LinkedIn. And Mentor Analytics also has a LinkedIn profile page. So I’m happy to be found on any of those sources. All right. Yeah. Well, thank you. Thank you for having me.
Kirill Eremenko: Thank you so much, Wesley, for coming on the show. Really, really appreciate it. And lots of knowledge bombs. I can’t believe how fast this hour flew by. Thank you so much. I’m sure lots of people will get value out of it.
Kirill Eremenko: So there you have it ladies and gentlemen, that was Wesley Engers, a top rated data scientist on the most popular marketplace for freelancers called Upwork. I hope you enjoyed this conversation as much as I did. And my personal favorite takeaway was the whole notion that data science is recognized as one of the fastest growing and one of the most important areas in freelance work right now. That is a big statement that not only data science is important inside companies in house, and as consultants, data science consultants are adding value to companies, but also that Upwork, the largest online marketplace with freelancers recognizes data science as a very critical area of its offering. That speaks volumes.
Kirill Eremenko: That means that there’s huge opportunities for those of us who want to be freelancers, who want to engage in freelance work, whether it is as a career choice and full-time work or whether it’s a part-time thing that you might want to do on the side while you’re pursuing other passions, other areas of your career, the opportunity is there. And then that’s important because as long as the opportunity is there, then it’s our choice whether to take it or not.
Kirill Eremenko: And of course, very huge thank you to Wesley for sharing all his skills and knowledge on this podcast and giving us some insights into what it’s like to be a data science freelancer. If you want to follow up with Wesley and connect with him, whether it’s to learn more and ask him more questions or whether it’s to hire him for a project, you can find all of the links to his LinkedIn and website and Upwork profile, which I highly recommend checking out even if you are… Like if you are starting out in freelancing or even if you’re just considering it, highly recommend checking out Wesley’s profile on Upwork. So all of those links will be available at the show notes, which are at www.superdatascience.com/323. That’s www.superdatascience.com/323.
Kirill Eremenko: And that’s also the link that you can send to any of your friends or colleagues who are considering becoming data science freelancers or who are already doing data science freelancing and need a bit of a boost or some tips and hacks to get better at it. Just send them the link www.superdatascience.com/323 and help them grow their careers as well.
Kirill Eremenko: And on that note, thank you so much once again for being here today. Really appreciate your time and I hope you enjoyed this podcast. I look forward to seeing you back here next time. Until then, happy analyzing.
