Data Science

Tom: Hello and welcome to episode Eleven of the Make Things Better podcast. Today I have Ben Atha, founder of the Developer Academy on the podcast, and I'm quite excited to talk to you about data scientists and data science in particular.

So thanks for coming on, Ben. How are you doing today?

Ben: Great. Thank you for having me. I'm really excited to have a chat with you about about data science and the industry.

Tom: Yeah, cheers for coming on. I appreciate your time.

Do you want to start off by telling us a little bit about the developer academy and how you started it?

Ben: Sure, so I started the business about three years ago. It kind of came out of having attended a lot of the Sheffield digital meetups especially geekbrekky.

When I was in a previous role, I was talking to various people there and it became quite obvious that there was a skills gap not just in Sheffield, but nationwide for developers, for coders. I can't code and I still can't code, but it kind of set me off down a path of well how do we help with this skills gap? How do we train people? And through various conversations, it became quite obvious that if you do want to learn to code, if you're interested in it, you can learn to code.

The other thing that really was helpful when I started the business was that the companies that I was talking to, the tech companies, were all saying the same thing. You know, 'we don't care about formal qualifications, all we care about is, can they code? Can they prove it with a portfolio?'

But what they really struggle with is people that have the soft skills to go alongside the coding skills. So can people problem solve? Can they explain something that is very complex in a simple way? Can they communicate what they're doing with all the stakeholders in the business? Can they learn themselves? Are they able to teach themselves because the technology sector is constantly changing. There's always a new framework or a new coding language, and have people got the passion for coding? Do they really enjoy it?

And they are the skills that they really, really, struggle to find. So that kind of set me off on a path where I started to think about my own situation. I mean, I was at that time in a job I didn't like, and I'd spent a long time in jobs I didn't like. I was just following the money a lot of the time. So I kind of realized, well, there must be other people in jobs that they don't like, they've probably just done the same as me. They've left school, maybe without an idea of what they wanted to do and have gone down a path of let's just follow the money, maybe sales jobs, marketing jobs whatever they are.

And through those jobs, they will have built up all these soft skills without necessarily realizing they're building them up sometimes. So I started to throw around different business ideas and came up with this idea of, well, there must be, because if you research like coding bootcamps, they're all full time. They're all 12 weeks, 14 weeks, or something like that, which is great. It's really quick. You know, you can learn to code and they'll help you find a job. But there must be a lot of people that can't afford to do that, can't afford to quit their job for three or four months, learn a new skill and then get a job.

So the idea we ended up with was a part time coding boot camp whereby people attend in the evenings two evenings a week. It takes double the time of a full time boot camp, so ours is over 24 weeks.

But people can retrain around their job, around their commitments in 24 weeks, you know, it's still a short period of time, six months. And then obviously we help them find a job and that's worked really well for the last three years.

So that's kind of where we're up to now. We've grown quite nicely over the three years. I took on a director of education to handle the education side of the business because like I say I can't code, we've now hired three full time instructors and we've got about 15 to 20 contractors who do the part time bootcamp. So they are all developers in the industry.

During the daytime, we simply pay them as a contractor in the evening to deliver the boot camp. So yeah, it's been going really well and that's kind of where we're up to now.

Tom: Yeah, that sounds really good. And I quite like how you can't code yourself because it kind of just gives like a different angle to it because it sounds to me like it is more about helping people develop another skill set alongside their job so that they can maybe have more flexibility and transfer and go into a different industry which they possibly enjoy more.

And they still got those skills, as you say, from like a sales job or marketing job, whatever it is that they've done before. I think that's dead interesting, really, whereas I feel like a lot of boot camps would be led by someone who maybe is just solely focused on the code. Well not solely, but you know, it may be just a bit more focused on the code. And so they might not have thought about those other considerations as much, perhaps? I'm not sure. Anyway, so one thing I noticed on your website recently was a blog about data science, and this is a term that's quite new to me, really. And so I wanted to ask you, what is data science, first of all?

Ben: Sure. So, data science is a field that uses scientific methods, processes, algorithms and different systems to extract and apply knowledge and actionable insights from data.

So, I mean, data is a big thing. It has been a big thing for a long time. Lots of different companies are generating various different amounts of data. So what a data scientist will do is collect that data and sort and organize that data so that they can pull out insights that the business can use.

So a good example could be, I don't know, maybe you've got a product that's on the market. You can start building up the data about what features people are using with that produce, what features people aren't using with that product.

And the actionable insights could be, well, you know, people aren't using this particular feature. We don't need that anymore or people prefer this feature over that one. So let's focus more on those features, that kind of stuff. So that's helping you with product development.

Another example is you could have machines in a factory, if you're able to extract simple data from that machine, so when the machine is on and when it's off, how long that machine is running for, how much electricity is that machine using throughout a 24 hour period?

You can start to understand, well, it's more cost effective to run that machine at nighttime then it is at daytime because the electricity is cheaper or it's actually more cost effective to leave the machine turned on overnight than it is to turn it off and then turn it on again because some huge machines take a lot of time to power up, so that is then helping you reduce your electricity bill.

There's other things you can do with machines as well, you can try and predict when the machine is going to break down so that you actually carry out a service on that machine before it breaks down because it is quicker and more cost effective to service the machine than it is to wait for it to break and have it out of action.

So I think that's the key with data science, it's taking data and creating valuable, actionable insights that have got commercial benefits to a business.

Tom: Yeah. And a bit like what you were saying throughout your course, a lot of that is probably to do with problem solving, right?

Ben: Absolutely, yeah. I mean, a massive focus on our course is we get the students to do a lot of presentations and often they present them to me because I don't understand what the hell they're doing.

So it's a really good exercise for them because there's no point telling me all the technical coding stuff that you went into. There's no point telling me 'Well, I've got the data set and we created this algorithm and the algorithm does this that and the other'

I'm going to switch off. I don't care about that. I'm running the business. What I need you to tell me is 'We got some data, we did this with the data in a really basic way, and this is what we found, this is the insight that we found. Don't do this, do this, do that, do that. Or if you do this, this will happen. If you do this, this will happen. If you don't do this, that will happen.'

So we get them to focus a lot on those kind of presentations, explaining complex things in a simple way and trying to get buy in from various stakeholders because that's what they're going to do in a job. If you go over to the marketing department and spend 20 minutes going on about algorithms and that kind of stuff they're going to switch off too, they want to know what do we need to change to make the ads better? Or what do we need to change to make more sales? So that's that's a big thing of what we get them to focus on on that course.

Tom: Yeah. Are there any common misconceptions about what a data scientist does?

Ben: I mean, don't get me wrong, you do need to understand maths. You need to have a good understanding of maths, probably A-level Maths is what we run on. But you don't need to be amazing at maths.

This is the thing. You know, you've got so many tools available now to help you in that side of what you're doing. And I think a lot of the time with data science a lot of it is about logical thinking and trying to think logically and also trying to problem solve.

That's the biggest thing. I think maybe some people think it's very much 'Oh, I need to be super smart, super clever and come up with a really cool algorithm and I need to use machine learning or I need to use AI on this.'

You know, that's a big thing, I think people often think data science is all about machine learning and AI, and that super complex, super difficult stuff that only people like professors and people at university are doing.

And don't get me wrong, there is still an element of that. You know, there is still a need for that. But you don't have to be super smart, super intelligent, super great with maths.

You just need a kind of basic knowledge really and you can you can build the rest out from there. So I think that's probably the biggest misconception.

Tom: Yeah. And how much code is involved in being a data scientist?

Ben: There's quite a lot. I mean, we focus on teaching Python as a codinglanguage, which is really accessible as a coding language it's probably one of the easiest coding languages to learn. So that's great because it makes the whole data science field far more accessible as well.

And yeah, there is a lot of coding involved. But there is packages you can use with Python and there are tools that you can use with Python to make the job easier, to make the job quicker.

nd it is all about speed a lot of the time. You know, cloud infrastructure now with AWS and Azure, a lot of it's already done for you and you don't have to create a brand new machine learning model.

A lot of the models are already there. You can take them and adapt them. So it's getting far more accessible and far easier than it probably used to be maybe five or ten years ago. Yeah, there is a lot of coding, but there are a lot of tools there to make it quicker and easier.

And I think, maybe another misconception with data science is that you don't have to spend two or three weeks putting together an amazing machine learning thing.

A lot of the time you just need to get yourself 70, 80% of the way there with some very simple coding, with some very simple pre-made tools, so you can very quickly see 'OK, is there a value in this data or is what I'm trying to do going to produce an outcome that's commercially viable?' You can probably do that in a day or two. And that's very quick. And you can go 'Right. No, that's not. Let's move on. I think that's maybe another misconception. So, yeah, there's a lot coding but a lot of the time you can be quite simple and straightforward with it before you get involved in the real big stuff.

Tom: Yeah, absolutely. That makes sense. And then other then like maths, which you already touched on a bit, what other skills are needed to be a data scientist?

Ben: So we teach about data, how to clean your data, how to organize your data. We do go into machine learning, so supervised and unsupervised machine learning. We cover big data. We cover cloud infrastructure, so AWS and Azure and how those tools can be used with data science. And I think they're probably the main skills, you know, we touch on algorithms as well, obviously that's part of the maths.

But, really it's having a bit of a broad knowledge of the tools that are being used within the industry. You know, you don't have to have a great in-depth knowledge to get into the industry. All the companies are different, some companies will just focus on AWS. Some of them will just focus, maybe on a Azure, for example, with the cloud.

So we touch on all of these different tools, you know, we touch on machine learning and all these other areas, just to give you an understanding and the knowledge of all the fundamentals of what you're trying to do and the fundamentals of all these available tools that you can use, I think that's really what you need to learn and what you need to know.

And then you need to get into the industry. You need to get that junior job where are you going to be working on real datasets, you're going to be doing real presentations. And yeah, I think the soft skills definitely, you know, presenting complex stuff in a simple way. You really need to hone in on that and also hone in on logical thinking. You know, here's a problem, I logically have to work that out, you know, how do we do that? And have the soft skills to take the time to look at what it is we're trying to achieve. And also commercial understanding.

Sometimes people can go 'I've pulled out this really great insight from the data. Look it's going to show me that' but there's no benefit to the business. I mean, it might be a really great thing, and it might be a really great thing to tell people about.

Oh, we found this in the data, and that's really interesting, but there is no commercial benefit. It's not going to save us time. It's not going to save us money. It's not going to help us build any new product. It's not going to help us in any way, shape or form. It's just an interesting thing. And I think, you know, being able to work out what those two are is quite important, and it can sound like a simple thing. But yeah, I think that's a big thing. Having a commercial understanding of what you're trying to achieve is really important.

Tom: Yeah. And maybe that's why people who already have like a decent commercial understanding from previous jobs who are coming to the developer academy, they they may benefit a lot from what having that past experience. I can definitely see how that kind of ties in with learning these skills and they complement each other really well. And you mentioned AWS and Azure there and I know that Hive have used Azure on our Department for Education projects, and I know we've used a AWS on, I think our Sheffield Mental Health Guide project, and I'm sure we've used both of them on many other projects as well. But I don't really know what either of those platforms do. I know very little about them. Could you tell me briefly, like, what do they do?

Ben: Sure, so they are cloud infrastructures, so they're both very, very similar. AWS is created by Amazon so it's Amazon Web Services, and Azure was created by Microsoft they both pretty much do the same thing. On a basic level, instead of having a server, so an old fashioned server sat somewhere, sat in your office, for example, with your website stored on there and maybe all your data and maybe your emails and whatever else. Instead it's stored on a server in a data center. So Amazon Web Services have a lot of data centers across the world.

So all of your information is stored on that cloud infrastructure, in a basic term. Things like Google Mail and everything, they all run on cloud infrastructure. But what they do as well is you can actually run, you can actually create applications in the cloud, so you can build machine learning modules in the cloud that uses that cloud infrastructure, so you're using AWS's power and infrastructure to run that model to make it work. So that's just the basic level.

It goes into great, great detail. There's a massive amount of things you can do. There's big data there as well, I mean you can imagine Facebook's got all of their stuff stored. There's a huge amount of big data that they've got stored. And AWS is used to store data, but also pull insights from data, analyze it, track it, keep it all secure, that kind of stuff. That's just kind of a basic overview.

Tom: Yeah, there's probably too much detail there for like one podcast alone.

Ben: So yeah, I mean, we say to our students, you know, AWS and Azure they go from zero to blow your mind really, really quickly. You know, it's very easy to create an account and get something up and running. But there's so much to it. I mean, if you look for AWS training courses, they have their own training course on it and there's seven or eight training courses that take years to complete. They're absolutely massive.

Tom: Yeah sure, that definitely helps me understand it better. So thanks a lot for that. And how would you recommend someone go about learning data science?

Ben: So there's a number of different ways you can do it. I wouldn't say any way is better than the other. Everyone has their own way of learning, and it's the same with learning to code. Everyone's got their own way of learning. There's a lot of different ways of doing it.

With data science, you know, you could go and get a formal qualification, you could go to college or university. You can go down that route. You can teach yourself. The great thing there is that if you went on YouTube, you've got hundreds and hundreds of videos where you could probably learn from YouTube.

There's websites like datacamp, you can learn everything on datacamp and you just do it all online as you go. There's lots of different forums you could go to. There are many different meetups throughout Sheffield or in your area, and you could try and talk to people and find out different information. You can learn that way.

Or you can do something like a boot camp, like ours. I mean, the good thing is you've got the kind of pluses and minuses of each one. So if you go down the college and university route, that's great. You get a formal qualification, but you're not really getting any job skills at the same time. And I think those institutions maybe have a bit of a bad reputation for teaching soft skills. They focus very much on the harder skills on the curriculum.

Teaching yourself is great. But, you might learn a lot of things you don't need to learn. That's a problem there, you haven't got any real guidance on well just learn this, this and this. There's no set course there and there's also nobody to ask questions to, and you can end up down a rabbit hole a lot of the time. So, you know, you can teach yourself for free online, but there's no one to ask questions to and you might learn the wrong stuff.

And then with the boot camps, ours is £5,000, so it's not cheap. Obviously, you can finance that with our finance partners, but the positives are it's instructor-led lessons. You've got that constant support. There's somewhere there for you to ask questions. And we've written the curriculum with industry, we've worked with companies like Twinkle in Sheffield, like Deeper than Blue.

They've helped us create the curriculum based on what they want, what they want to see. So you know that when you do a bootcamp like ours that you're learning the skills that companies want, so that you should be able to very easily get a job at the end of the course.

So, it just depends on how you learn really and how you can do it and people have done in different ways, you know, I've come across lots of people who have self-taught themselves. Lots of people who have gone to uni and lots of people who have done boot camps. It depends how you learn really.

Tom: Yeah, make sense and actually I looked on your website. A few of the things that people really liked about your courses was how they just have that structure so that they're not like maybe procrastinating, and then they can be flexible about that structure as well.

But just to have that structure, I think that's really useful when you're learning anything and just having that person to ask questions to, because I know when you're learning something difficult, you can easily get stuck and that can be quite off-putting and challenging. And you know, it's hard to regain focus sometimes if you do get stuck on a problem and it's taking you a while to overcome it. So I can see how that can be really valuable.

So where can people find yourself then or the developer academy on social media and what is the name of your website as well for anyone who could be interested in one of your courses?

Ben: So the website is thedeveloperacademy.com and you find all the information about the courses we've got.

Our social media accounts are

Twitter: @TheDevAcademy LinkedIn: www.linkedin.com/in/ben-atha/ Instagram: thedeveloperacademy Facebook: thedeveloperacademy

We're regularly updating our social media about what bootcamps we've got running and what students learn on the boot camps. We do put out a lot of information and a lot of helpful information. You know, you should learn this or you should learn that. This is how to do this, this is how to do that so you can pick up a lot of free stuff from us as well.

We often run free courses throughout the year so you can come along and learn a little bit of how to code and that kind of stuff. So, yeah, check out the website, check out the social media accounts.

Tom: Yeah, awesome, thanks a lot for coming on Ben. I really appreciate it. I've learned so much in just 30 minutes so cheers for that.

Ben: You're welcome. Thanks for having me. It's been a pleasure talking to you.

Tom: You too. And I hope you have enjoyed listening/reading this, and I hope you have a great rest of your day.