Hello Patrick, great to get a hold on you. Tell us about yourself, what made you join Parashift and what you do here.

Hi, my name is Patrick and I have been working at Parashift for about 7 months.

Parashift sparked my interest for a multitude of reasons. First of all, I had multiple customer projects in my past where document extraction was a huge pain. And at the time I could not believe that this problem still exists. During my bachelor studies, I also discovered my interest in machine learning, which we deal with quite a bit here at Parashift. Last but not least, I was looking to get back into an office after working remotely for quite some time. The last point obviously did not go as planned, due to an unnamed, global pandemic.

I was hired to help out the machine learning team with software engineering and to build a bridge between them and the backend developers. Right now, I am doing a bunch of things like help designing and implementing a new microservice architecture, build a python client for our REST API and construct libraries and tools to make the life of other software engineers easier. When nobody is watching, I also like to do some machine learning on the side.

What would you say is your superpower?

I don’t have one. Boring, I know.

What would you say is currently the most challenging problem you’re working on and why?

Recently, we made a few changes in the way we build software in the ML team. We built new libraries, made architectural changes, and revamped the way we package and deploy software. While I like to think that these improvements help with stability and scalability, we also realized that they added new complexities to how we work. So, my goal right now is to get rid of these complexities so that everyone can be as productive as they’d like to be.

If you could switch your job with anyone else within Parashift, whose job would you want?

That’s a good question. To be honest, I am already doing what I want to do. Otherwise, I would be looking for another job… However, if it meant to also magically gain the required skill set for the job, then I would like to switch brains with one of the machine learning engineers. I always struggled with math, so this might be the perfect opportunity to figure out what I am doing wrong.

What does a paradigm shift mean to you and why did you decide to join a crazy bunch and contribute to one yourself?

To me, a paradigm shift means to rethink established processes from first principles and possibly come up with a better solution. And I think this captures the way we do things at Parashift extremely well. Not just for our product but also for how we work. That is why I am here.

How do you envision the next five years for Parashift?

Parashift has an interesting mix of engineering and research challenges in front of itself. I am looking forward to seeing how we get better and better at customer use cases by iteratively building a better product and applying the newest research in the field.

Before we wrap up, one last question. For people’s next travels after COVID-19, what inspiring book or podcast recommendations would you give and why?

Oh, there are too many and they vary a lot by who is asking me. When it comes to books a few of my favorites are:

  • Masters of Doom
    • If you like video games, then this book needs no further introduction.
  • How to live a good life
    • To live a meaningful life, one needs to define meaningful. This book describes how stoic philosophy is surprisingly applicable today and might be a viable philosophy of life.
  • The Soul of a new machine
    • A story about hard-working engineers who build computers, not for the money, but because they love it.

Again, my podcast list is long. I like to subscribe to many and listen to a few. Here are some I enjoyed recently:

  • Lex Fridman Podcast
    • If you like machine learning, robotics, computer science, … then you can’t go wrong here. My highlights are the episodes with Chris Lattner and Jim Keller.
  • The Handmade Network Podcast
    • In an area in which we rely more and more on abstractions, we might forget how computers work. And even though computers are faster than ever, software is getting worse and that’s something we should at least think about.
  • Hardcore History
    • Podcast is probably the wrong description and an understatement. Just listen to Blueprint for Armageddon and don’t be surprised if you travel 20 hours into the future.

Great one Patrick! Very much appreciate you taking the time and sharing insights and some wonderful resources.