We are here today talking with Dan Stein about his recent change from a completely different successful career; to start his journey as a Machine Learning Engineer. Hopefully his story, although an unusual one; can help and inspire people coming from all walks of life who might want to make a similarly challenging career move and get into the space…

So Dan, for those who don’t already know your alias, would you mind giving us a quick background on your career before Machine Learning?

I am best known as Brit-nominated multi-million record selling music producer ‘DJ Fresh’. I’ve had three number one and ten UK top ten records and produced for artists ranging from The Petshop Boys to Diplo. I also founded the independent record label Breakbeat Kaos and early 2000s web forum www.Dogsonacid.com, among other things…I’m a bit of a workaholic!

What motivated you to choose Machine Learning as the next step in your career?

I started programming when I was about 7 or 8 messing around on my dads old Atari, I was always fascinated with computers, but their application was limited at that time, and music seemed more exciting! I went to Kingston University studying Business and IT, set up the formative drum and bass band ‘Bad Company’ and soon after started the award-winning web forum DogsOnAcid.com. At that point, we had 30 million unique page views a month. This was before Facebook etc. so it was a very different and markedly less exciting internet. Success with my record label Breakbeat Kaos which had recently signed and brought over to the UK Pendulum and later discovered artists like Chase & Status, Nero and Sigma led me away from tech for a while. Though I was briefly involved in an IP protection startup called Darkstar with Chris Parry (XFM, Fiction) and the hacker/author ‘Chris McNab’ among others. Darkstar protected music and media on the early P2P networks, and had clients including Buena Vista and EMI but unfortunately, it was a tough time for the labels and it was hard to secure the budgets necessary to empty over 30 coders around the world and a building full of servers.

How did you begin to make the move?

I’ve always loved programming, it’s a creative art particularly with the strides forward in browsers and modern UI and interfaces and while it has foundations in maths and science, the palette is infinite, you can create almost anything you can imagine which appeals to my creative side. I also felt that networks like Spotify were changing the way that listeners interacted with music and for me to understand this new landscape I needed to make some big changes.

In 2016 I had cancer, which thankfully I’ve now recovered from, but I guess it pushed me to make the changes that had been eluding me for so long. I also recently got married, and never liked the idea of being on the music front-line forever, it’s a tough gig (excuse the pun). I’ve always been a fan of people who had multiple careers like John Glen the astronaut who became a US Senator, I don’t believe that life is one narrow path you are confined to, it’s a matter of mindset.

How have you found the transition?

Although I’m famous for being a workaholic (lunatic?) with those that have worked with me, I’ve never worked harder than I have in the last couple of years. I’ve always studied in my free time, I contributed to a paper a couple of years ago with my friend the author John Powell about the physiological effects of music and have always kept abreast of things in the tech world. I’ve always believed to really understand a business you need to start from the ground up to understand all the components. When someone tells you something is impossible, you need to know if they’re wrong and why.

In order to get my head back in the game, I went on a live-in boot camp WeGotCoders, founded by Dan Garland and taught by Ben Pirt. It was a Ruby course, which wasn’t ideal, as I wanted to focus on JS and React, but the setting was like something out of an old ghost story. An enormous mansion in Hoddesdon with an underground aquifer in the basement, it was a trip! I met some great people, some who were also making the transition from quite different backgrounds, one was a professional dominatrix! A lovely Canadian girl who was a carpenter. Both now have good development jobs in the Ruby on Rails world, Dan and Ben were fantastic teachers.

What hurdles or objections have you faced to get where you are now? How did you overcome them?

It’s hard to transition to something new especially if you’re used to having earned respect and the benefit of the doubt in a different field. I know it’s been an eyebrow-raiser for people. I always think to myself, five years from now it will all be clear. I’m really excited about the future. But I was really lucky to find HT2, I was aware of what they were doing as I live nearby in Oxford and they’re a Mecca in the online learning world. I found a fantastic recruiter, Carl Weaver who helped make the introduction, and it just felt right. My wife and I are planning to start a family so having a ‘day job’ is just what I need for the next five years or so.

With no recently earned Data Science credentials, I had to prove myself. I started off in a React-Redux (JavaScript) Developer role. I spent a lot of free time working on an idea I had for a potentially game-changing ML platform. I proved to the guys at HT2 that I could dramatically improve their ML workflow and accuracy, and when their Data Scientist/ML Developer left to find something more local they offered me a role that was just what I needed. We are now employing the platform that I’ve built with the guys at HT2 and think it could change the landscape of ML in the e-learning world.

What is the reception you receive from colleagues and others in the industry about your story?

Confusion, shock, disbelief!

I’ve had a few people (some of whom you might have heard of) say to me ‘Dan if it was anyone else I’d be surprised’. I guess I’ve always liked to push boundaries.

Obviously, fans of my music want me to make more, and I am, but right now I see ML and coding as a fresh palette and something that I can potentially make even bigger waves with than I have with music. One day I might be able to transfer what I am learning about ML and artificial intelligence and use it do something with streaming platforms to help artists have more control of their music.

Do you have a career highlight in Machine Learning so far?

My first project as ML Engineer has been SID for Curatr, HT2 Labs LXP learning experience platform, we just won an award for SID from Brandon Hall (one of three that HT2 took away that evening) and it’s been shortlisted for an Owen Mumsford Innovation Award.

Has it been difficult to essentially start from the bottom of a whole new career after being at the top of your game in another?

It’s been sobering, I was lucky to find a great home. It was hard understandably winning the guys over at the beginning and the first six months were tough, but they were really supportive and ultimately gave me their blessing by promoting me and encouraging me to bloom in an area that I am incredibly passionate about. I’m lucky to have built some foundations over the years with music to have met an incredible woman I’m lucky to call my wife and to have calmed down enough to be able to fit into an office environment. It’s been an almost spiritual change of perspective to turn down financially rewarding gigs to almost start from the beginning again, but sometimes you have to think past the short-term and make sacrifices for the long-term.

I’ve been making hard decisions like this since I was in my teens, I think it’s easier to do that when you come from a world as competitive as music. It’s important not to let your life be ruled by fear. I had a similar reaction when I wound down a large part of my record label activities in the late 2000s, but in retrospect, it was the right move. I’ve always said about business, knowing when to get in is just as important as knowing when to get out!

There was one day early on at HT2 at an away day where after being there for a month, one of the guys, slightly drunk at dinner burst out with “We just can’t believe you’re working here, it’s so weird but we didn’t want to say anything”, it was hilarious! I’ve made some good friends there and it’s fun being able to tell them the crazy stories about ‘my other life’. They’re a great bunch and they do important work. Their Founders Alan and Ben are very inspiring, Ben is a really inspiring speaker and a bit of a star in the e-learning world.

What do you find rewarding about your career now?

It’s great working on a cause that is truly positive at a time when learning is becoming more and more important in our daily lives and accessible in so many ways. I’m a bit of a geek! I love to figure out how things work. My work with machine learning echoes my experiments with synthesis and finding patterns in music, I’ve always had a very scientific approach, which is why drum and bass particularly appealed to me so much. It’s been great working closely with talented people like HT2’s Data Scientist Janet Effron, and CTO James Mullaney. It’s very collaborative, and that’s something that has always been important to me.

In a nutshell, if you want to make an omelette…

Where do you aspire to take your Machine Learning career in the future?

I’d like to start by getting it to successfully predict the end of my last sentence.

Seriously though I really appreciate the opportunity the guys at HT2 have given me, and I’m very happy here so while I do a bit of growing up and hopefully having kids I intend to do great work for them and hopefully help make some advances in the field. Having this output also gives me the freedom to only release music that I’m really excited about. After that, who knows, I’m an idea machine, I have no shortage of ideas for Machine Learning, apps and new start-ups. But right now I’m enjoying the first time in 20 years I’ve had weekends with my family and friends, and having time to learn and do the things that interest me.

Do you have any tips or advice for others in a similar boat wanting to make a big change?

It’s ironic because people have always asked me that question about getting into music, another very difficult transition (from anything!). All I can say is that anything in life is possible if you don’t take no for an answer, believe in yourself and work really hard. I probably need some advice on having a balanced life, I am definitely in the workaholic arena, so I appreciate not everybody will be prepared to make the sacrifices I do. But even with balance, I think you can achieve things in the future you wouldn’t believe today.

Admittedly being a very technical Music Producer/Engineer it’s not such an extreme jump and I’ve heard lots of musicians moving into coding. The creator of React Native Dom used to be a musician and even Kanye West recently announced he wants to get back into coding. But with vastly more demand for developers (especially Data Scientists) than there is supply, we are in a time of transition. Some people are calling it the third industrial revolution.

If you want to become part of the future you have to bite the bullet and let go of the past. There are incredible boot camps where with a bit of dedication you can get into a role in tech in the space of less than a year and the old cliches about life as a coder and in IT couldn’t be more unrealistic. If you’re a designer you could get into UI design, if you’re a Store Manager you could get into running a web store. The internet is just a virtual version of our world, one day there will be jobs to involve and suit everybody that involve technology and the internet.

Without coming through the ‘traditional’ education route at a young age, are there any key skills or experiences you found particularly necessary to get into this space?

I have had some experience in academic data science through my time at Kingston University though that was a long time ago, I’d say ultimately my work ethic. Right now I’m studying data science, multivariate calculus, and three different languages. I read an insane amount of books, I spend my life doing online courses and trying to absorb as much as I can on a variety of subjects I’m interested in, while working a full-time job and still running a record label and an automated Spotify playlist I created for the 60 or so biggest labels in the drum and bass scene in the background.

Work doesn’t have to be tedious if you do what you ’re interested in. If you’re reading this and you don’t truly love technology and computers, then you should be trying to figure out what you DO love (or maybe check out the tune ‘Throw’ from my old album Escape From Planet Monday). I know a guy who’s a really successful and happy gardener who used to be a really successful (and much wealthier) banker. Follow your passions, follow your dreams, we’re only here once! (until ML disproves it).

Do you have any advice for anyone tackling interviews and application processes from your experience?

Wow. Interviews in tech are hard! I spend a lot of time on Stack Overflow answering questions, that’s a great way to practice. I don’t really agree with the current technical interview format. As someone who’s run a successful tech company (albeit years ago), I look for people who have the drive, ambition and that will fit well into my team. You can see people’s real work on GitHub and Stack Overflow. There’s nothing wrong with a few questions, but just like with university and school, being great at tests doesn’t mean being a great member of a team or someone who might come up with fresh ideas for your company. If you find yourself in an interview where they care more about your ability to answer the kind of question you’re never likely to face without the internet and time to figure it out, ask yourself, is this really a place I’d want to work?

Is there anyone in particular who inspires you in Machine Learning?

Not just Machine Learning but the JavaScript and React community in general, people like Sofie Bits who just left React Core, Michael Jackson, creator of React Router, Dan Abramov creator of Redux (now at Facebook). The Tensorflow core team who I’ve been lucky to get to interface with people like Shanquing Cai, Vince Mueller, Nikhil Thorat, and famous persons wise Ray Kurtzweil who has also straddled music and technology, I’ve been using his studio gear for years!