I recently had the pleasure of interviewing Nic Ryan, an established Data Scientist, with a wealth of knowledge and experience, and a huge following on LinkedIn.

He was kind enough to share his personal story of how he entered the field and what inspires and motivates him to help others.

An inspirational story full of insightful tips and advice.

You have a large following on LinkedIn and are such a positive and helpful influence in the Data Science community, always happy to offer advice and support to aspiring and junior Data Scientists. What’s your motivation?

Oh wow! That’s kind of you! I guess a couple of things really. I live in Bundaberg in Central Queensland so by location I am isolated.

I am the only Data Scientist/ Data Analyst / Dashboard Builder/ All Round Data Guy at my company. So, professionally it is isolating. The nearest data scientist is in Brisbane 5 hours drive away. So, talking data science with real people who know things is amazing. Yeah, nobody is doing data science in Bundaberg.

I have worked in different roles in different industries for many different companies, I’ve led teams and I’ve worked autonomously. I have been doing this stuff for a fair while, if I can share a couple of insights with people then that’s awesome.

I remember how tough it was getting that first job in an Actuarial Department. Well, I took a year off to go surfing after uni, so that wasn’t the best career move but there was amazing surf from Bells Beach all the way to Bundaberg. Even at 22 years of age I loved Bundaberg but the time wasn’t right.

Getting that first start is hard, even though I couldn’t get a job in the field I wanted I just took a job, any job I could get. My wife and I had moved to Sydney without work, we had limited money, not even enough money to buy bread, so we made our own. Looking back on it they were good times, even though it wasn’t always fun. Anyway, soon I was able to move to other things. I would advise others to do the same, working close to data but not as a data scientist is better than not working at all, trust me! There’s the money side of things but there’s also the confidence/ mental side of it too.

Elliot Heads a beautiful spot next suburb to where I live

How did you find the transition from Actuarial analysis to Data Science? It obviously set you up well with a strong background in quantitative analysis, but what were some of the challenges you faced?

You know what? Here’s the thing, I never really was keen on actuarial studies. I just didn’t ever really like it, most of it back in the day was done on spreadsheets, and ok there was a bit of math and stats, but it just wasn’t fun to me. I naturally gravitated towards the stuff I found more interesting at uni like stats and probability and ended up building insurance pricing models. However even at uni I knew I wasn’t keen on actuarial studies. Yet, I still saw it through. That was tough. I should have really swapped to computer science and stats, someone recommended it, but I didn’t listen at the time. A time when grit and determination worked against me.

Back in the day in actuarial departments it was mainly SAS that you’d use in actuarial pricing, well actually overuse. SAS was used were SQL should have been for data manipulation and munging where really a scripting language like R or Python would have been much better.

The main challenge for me was code. I didn’t really know any code, I have had to teach myself. Each person approaching data science will have something they aren’t strong with:

  • Business acumen and domain expertise
  • Math and stats
  • Code

MOOCs were my friend, that and dedicating 2-3 hours to my craft every day.

What prompted the transition? Was it natural or a deliberate decision to change careers?

My oldest daughter Auralie was unexpected, I was a young dad at 26 years old. Funny as I write this because my daughters Auralie and Lucia are my best buddies, we do everything together from basketball, surfing, skateboarding, go out for bubble tea together etc. They are awesome but finding out about Auralie at the time was scary as hell. Single income, Sydney is an expensive city, not cool.

I wanted to be around for my kids, and I didn’t want to do actuarial studies (I was on the last part of my actuarial exams at this stage), so I moved back to the Central Coast of NSW where I grew up, it’s about 100km north of Sydney.

I saw a job at a bank for a credit risk analyst in Sydney and decided to give it a spin. Believe it or not it was quite interesting, still using SAS and a bit of SQL. They were hard years because I had an epic 2 hour each way commute, so 4 hours per day of just dead time. I was getting up at about 4.30am to get to work a bit early, so I could leave a bit early and still get home to see the kids maybe at 6.30pm.

To pass the time I started doing MOOCs on Coursera on the train. Well, that was it! I was completely hooked, and I loved it! I couldn’t understand why R or Python wasn’t being used at work, nor why we weren’t using some different techniques I was learning. I built my skills up massively. That train trip from Woy Woy to Sydney made me, it is kind of funny that it could have been an awful experience, on paper it was but I am so thankful for it. It was like a master’s in data science.

I continue to spend 2-3 hours per day on either learning dev or data science.

Me with all my girls (Amber not shown)

What was your inspiration for starting DataFriends? What’s its main aim?

I think for me working for a big corporate like a bank or insurance company is just not a natural state. That’s why I have gravitated towards start-ups. You tend to have more responsibility, wear a few different hats and you just learn a lot more. I tend to get bored easily, banks move oh so very slowly and it kind of used to kill me at times, so DataFriends was initially about consulting and mixing it up for me. I got a few gigs doing consulting work and it was amazing fun.

Then I realised that nobody knew what the heck a data scientist was. So, it also became about education, I have had the opportunity to speak at a few places about data science and just what the heck it is. So that was fun. I always loved mentoring newer data scientists anywhere I have worked, so it probably comes from there too.

The problem with DataFriends was a recruiter saw it, mentioned they wanted a Head of Data and Credit who had domain expertise, could talk and was also “hands on”. They’d tried unsuccessfully to hire for 3 years but were interested in talking to me. A plane trip later I was talking to the CEO and I got offered a job – well Senior Data Scientist job because they thought I was too young for the top job. But after they couldn’t find anyone else they gave me the top job. The figure they offered me looked like a phone number, so I took it, otherwise I would have just started consulting full time.

So, there has been a bit of a hiatus for the last few years on the DataFriends front, but I’m launching it again. Well, my wife built the site, because she is amazing and understands design and things.


You’ve also assist your wife with the charity Coding with Grace, that she founded. Can you tell us a bit about it, and what have you achieved through it?

Where I live has about the highest unemployment rate in the country. There is work mining and building and in the trades for guys, but not much besides retail and tourism, cafes etc for women.

My wife saw that I was working remotely from Bundaberg for a company in Melbourne. “If Nic can do it I suppose anyone can do it!”. She had a look at remote working tech jobs and saw that most of the remote working jobs were for front end developers. So, she said this is the way forward for the region, the idea is to give the ladies a bit of a view of html, CSS and JavaScript, git and command line so that if they like they could do a bit more study on their own and get a job as a front-end developer, designer or whatever. We haven’t had any support at all from anyone, not council or big business or anything, a few of the local schools donated old laptops which was amazing. We don’t charge a cent to the ladies, so we don’t make any money on it. We ran a couple of 10-week courses last year with 2 great groups of ladies. We meet at cafes, or even just online. It’s like a support group with code, obviously a tribute to the legendary Grace Hopper.

Kind of funny but I am no front-end developer, I can google and hack my way around a website, but it isn’t really my passion. One of the ladies in the first course was an absolute gun. She had done some dev work before, more CSS and design, and man she was a star and passionate about it all! So, I just said to her “hey you’re awesome, you want to be involved?” and so she did a fair few of the lessons for the second course which was superb. She is now building an app to help manage her own and other people’s Endometriosis. I couldn’t be happier, she is amazing and of course I have given her my open-ended support. Some of the other ladies have gone on to further study in CS.

You’re active in a number of side projects, but which one are you most proud of?

Besides ManageEndo (thumbs up and applause to the amazing Morgahna Godwin) and Code with Grace, well there is something.

Here’s my dream, and I am going to be straight out level with you – I would love to create a digital bank driven entirely off bank statement data, I’d love to do it from Bundaberg and I’d love to do it on my own without funding or venture capital. It is just such a crazy story that it would be fun to make it work.

Everything I do, every little side project is just building a skill or skills in whatever I feel is missing in my skillset like hosting, deployment, html/CSS with this project in mind. Obviously, I have prototyped the categorisation engine, but it is the other stuff that until recently I had no idea about that excites me. If the side projects make money great, if not I don’t really care.

So, that’s the dream. Hard to do with all the other stuff on my plate, but really skills I build across everything compliments this epic project. I don’t have money to pay 10 people to bring this together for me, but say I eventually get some more dosh in the bank and a bit of runway I’d love to give it a go myself. As the sole income earner of the family it is kind of hard to do though.

I am still working at all the skills I need to make this happen. It’s a dream, it may not ever happen but who cares? It’s that North star I need to guide my efforts and dig in.



Some of the basketball crew

What does a typical day look like for you?

I sold my place on the NSW Central Coast and bought a place by the beach in Bundaberg. Houses are crazy cheap here, I love the area because the weather is like Hawaii all year around. I get cold when it gets below 24 degrees Celsius (75 Fahrenheit) and I reach for a cardigan. I love it here, the weather is great and the people are awesome. The guys I play basketball with and surf with are amazing, like give you their last dollar kind of guys. They are just regular kind of dudes. They keep me grounded.

Since I work remotely for a company in Melbourne, I have a bit of flexibility in my day. For whatever reason (I guess having more to do with the business than perhaps the devs do) I tend to stick to a 9-5 work day.

As the only data dude on the team, I tend to get involved in a lot of varied stuff, so for example today:

  • I presented a prototype for a funding model which predicts which invoices we are likely to see from clients in the next month. It was a prototype because it took me 3 days to develop. I am a big fan of rapid prototypes. Got some feedback, made some notes.
  • One of the guys from the business wanted an ad hoc report run on Supply Chain Finance numbers, just showing Key Performance Indicators. Data is on AWS, restoring local SQL Server DB etc, so a bit painful but pulled it all together for him.
  • There was a team meeting.
  • I manage the Power BI dashboards that we have as well so that required a bit of tinkering.

I have at various times updated the website or put together case studies. I am not precious about what hits my desk, I’m just a spare pair of hands. You must on a small team. I have also helped with Google Analytics, Adwords etc just because there hasn’t really been anyone else to do it. With my credit risk experience, I helped last week to find a new provider of credit bureau reports for us to use. So, yeah, it’s all varied work.

So, this is kind of how my day runs:

  • 530am: my little pug Amber wakes me up, I will have a bit of coffee and we will all take her for a walk at about 6am. The surf is across the road, I’ll check it out, see what the tide is doing and probably paddle out at about 630am.
  • 730-830am I will usually do a gym session either with my wife or one of my buddies. A mate of mine is training for the army special forces, he wanted me as a training buddy. I am an idiot, we’ve only just started and it’s going to be 20 weeks of hell.
  • 9am-12pm work, could be meetings, reporting, some model development.
  • 12-1pm lunch, I have two daddy-daughter days with my girls where I take them out wherever they want. My wife home schools our girls, I teach them code and data science outside my work hours.
  • 1-5pm work
  • I play basketball twice a week, so there’s a fair chance of a game being on around 6-7pm
  • I will always read to my girls, play board games or cards with them. I don’t really enforce a bed time, so they are free to read or sleep in bed from say 8pm.
  • 8-10pm I will spend a couple of hours each night just learning, or doing MOOCs, trying stuff out etc.

Surfing at the break outside my house

Amber the cutest pug ever

Where the magic happens (a pink table my daughters no longer use)

When starting a new project, do you follow a specific framework, such as CRISP-DM or something similar?

There are a couple that I have in mind, but really things move rapidly in the start-up world compared to my past lives in banks. Like I was saying I knocked out a prototype model in 3 days, in a bank you’d need 3 weeks just to get the 40 signatures on the paper to commence the project. I’ve seen model developments go for 18 months, only to realise that the original data they cut wasn’t relevant any more. This is a serious problem. The company I work for only has 10 people too.

The main ideas I have in my head really are Jeff Leek’s steps in a data analysis:

Something Andrew Ng has said before that he too believes in rapid prototyping, he tends to see people spend too much rather than too little time on a first cut rapid prototype.

I believe as well in putting together a simple heuristic, like rolling up everything by month and then doing a dumb prediction based on previous month or something. Say you are off my 30% here, well that’s a nice simple heuristic to compare your models to. If a spreadsheet approach beats your model, then you probably wouldn’t want to implement it.

About to go surfing, beach is just outside my gate

Disclaimer: The opinions expressed by Nic represent his own personal views and not those of his employer.

Leave a Reply