2018: Year in Review

I did a review like this in – 2017

It’s fun to look back on what I’ve learned and what I’ve accomplished. This will be contracting focused and side projects focused.

(Appologies to Julia Evans, I basically just stole her format)

I got married!

I won’t embarrass my wife online. But this year, after a stressful year of organising I got married to my long-suffering partner.

Here’s a photo of me rocking a kilt on my wedding day waiting for my beautiful wife-to-be to arrive.



started a business!

I started my own contracting business in 2018. This started around August 2018, shortly after I got married. As part of this I also launched a content business – www.probabilisticprogrammingprimer.com 

The content business has been successful, in fact some months I’ve made enough revenue to live on, and I’m really grateful to everyone’s who’s supported that work. I hope the course has helped you. I always thought that it was impossible to make anywhere near as much money teaching people useful things as I can as a software developer, and now I think that’s not true. I don’t think that I’d want to make that switch (I like working as a programmer!), but now I actually think that if I was serious about it and was interested in working on my business skills, I could probably make it work.

I don’t really know what’s next, but I plan to write at least one course next year. I learned a few things about business this year, mainly from:


In terms of contracting work, I’ve been working with startups and established companies. Some work I’ve done as an independent consultant

  • Designing and delivering a data strategy for a large fintech. Such as who they should hire, what data they should collect and what machine learning systems they could build
  • Data Science Dev Ops for a tier one bank. I delivered the set up of an analytics platform – largely based on jupyter at a large bank. Learned a ton about communication and the challenges of delivering results in an enterprise.
  • Data Engineering work, enhancing data quality and learning tons about using AWS Lambdas and Kinesis in data pipelines.

If you’re based in Europe (I’m open to remote work from the US too, but timezones might be tricky) and are interested in help with machine learning, data strategy, Bayesian Statistics or data engineering reach out to me at peadarcoyle[AT]gmail[dot]com


I left Zopa this year, which was the best environment I’ve found so far for doing machine learning type research. It was a hard decision and came after quite a lot of organisational change. Since then I’ve been doing my own thing.

I learned a ton there, about delivering on time, about the challenges of innovative projects and about mentoring and growing juniors.

I wrote this article on what it means to be a Senior Data Scientist as part of me figuring out the boundaries my increasingly leadership focused role. I got some great feedback from this, including a noted Bay Area company looking to interview me on the back of it. Shows the power of blogging 🙂

One thing I did at Zopa, that was super useful, and was encouraged by my leadership coach (reach out to me if you want a reference to a super dooper one based in London and engineering focused) – was I did ‘team contracting’.

A great post is here. But basically, it was a good way to ask standardised questions and figure out what was working successfully and what wasn’t. I learned for example that some juniors needed more hands on guidance, I learned that some needed some more code review and I learned that we needed to adjust the level of some of our internal talks. Exercises like this are super useful. And I recommend you doing them with your manager. Another great resource on this is help I have a manager.


I gave a few talks in 2018:

  • Modern Bayesian Workflow – PyData Meetup London
  • I did a similar version of the above talk at Signal Media – thanks to Miguel for the invite to speak to his awesome research team.

This year I didn’t have tons of time to do talks, largely due to wedding commitments and working on my online course. Next year I’d like to continue to do some talks on PyMC3/PyMC4 but I’d also like to expand a bit into more data engineering focused talks. We’ll see though 🙂


I also experimented a bit with a new format: the podcast! These were super fun, and I’d be open to doing more. I did the DataCamp podcast twice as a guest. I’d be interested in doing others 🙂

what I learned about doing podcasts:

  • It’s really important to give the hosts a list of good questions to ask, and to be prepared to give good answers to those questions! I’m not a super polished podcast guest.
  • you need a good microphone. I went out of my way to buy a good microphone.

Please reach out to me if you’d like me on your podcast. Super happy to talk about anything you’ve seen me blog about 🙂

blog posts!

I wrote a few blog posts in 2018. A couple of my favourites are ML Hipster trap , 3 reasons to learn Bayesian Stats, and What does it mean to be a Senior Data Scientist.

There were basically 3 themes in blogging for 2018:

pymc3/pymc4 summit

I was super excited to be part of the PyMC3/PyMC4 summit. We met in London, I got to meet Austin, Colin, and others face to face for the first time. I’d met a few of them. I learned tons about Tensorflow.

One thing I learned is that, it’s super valuable to get remote teams together face to face. Face to face is high bandwidth communication. Thanks to Google for all the support!


some things that worked in 2018:

  • understanding the boundaries of your role and setting them
  • doing open source work and even getting supported to goto the summit
  • starting a business
  • writing course content is very time consuming but I feel happy about the time I spent on that
  • blogging is always great
  • learning to communicate during the challenging parts of R and D work is super important. Not everything goes to plan but that’s ok.
  • learning to say no to projects. Not all projects are worth your time.
  • getting a leadership coach was super useful, as were conversations with some leaders in our field – machine learning is super new, but the lessons from other parts of technology are still valid.

Leave a Reply

Your email address will not be published.