Data Scientist Spotlight: Amanda Schierz
“I live for data. I love the data.”
Meet Amanda Schierz, a Data Scientist at DataRobot who is based in the UK. Before joining the team in 2015, Amanda combined her love of data science and biology into cancer research as a Computational Biologist at the Institute of Cancer Research in London. We spoke with Amanda about how she came to join DataRobot, obstacles she overcame throughout her career, and how automation empowers her to remain creative and passionate.
What is a typical workday like at DataRobot?
I focus on the development of novel tools and technique that enable our customers to monitor and manage their data and predictive models. We design and engineer tools that inform the customer of anomalous or drifting data and whether their models would benefit from retraining. I now focus on post-predictive modeling rather than pre-predictive modeling.
Is your work mostly with data scientists? Or non-technical users?
My main area of expertise was with biology, chemistry, life sciences, and pharmaceuticals. The only time now that I work with non-technical people is when they need domain-specific expertise on how to use DataRobot for the drug discovery process.
However, DataRobot is such a versatile tool that it doesn’t matter if you’re a bank or in pharmaceuticals – we have a solution for both.
What’s your background, and how did you end up at DataRobot?
I was a very geeky child. When I was 13 years old, I got the first ever home computer for my birthday. All computer programs were on tape, just normal audio cassettes. I taught myself, at 13, the BBC BASIC programming language. While all of my friends were learning how to put on makeup, I’m like, “I’ve built Space Invaders!”
In the 1980s, I decided to get a degree in Computing and did my dissertation on Neural Networks. I then did a Masters degree in Intelligent Systems and based on the dissertation, I was offered a full scholarship for a Doctorate, where I read text mining. I was one of the first people who did a PhD in modern text mining. With all the tools and techniques available now, my PhD could probably be completed in a month!
For my first post-doc, I worked with a man called Professor Ross D. King. He had just designed a robot scientist that scanned yeast to try and predict the function of genes. Through my work with Professor King, I learned about the drug discovery process. I ended up at the Royal Marsden, a royal-appointed cancer hospital, working on personalized medicine and developing new drugs as alternatives to chemotherapy. After a few years, I started getting sad because these drugs are only given to people when all else fails — it’s their last hope. The problem is that cancer is clever and it develops drug resistance.
Just as I was thinking about leaving the cancer research hospital, Xavier Conort contacted me from DataRobot. It was just one of those serendipitous moments. I had cyber-met Xavier through Kaggle and as DataRobot was expanding to the UK, I was the obvious choice for the role (I was the top UK data scientist and the Kaggle world leading woman for about five years).
What were the early days at DataRobot like?
At the beginning, we all did a bit of everything. I was doing QA, sales, engineering — whatever was needed. It’s been a great experience seeing the company expand.
As a data scientist, are you pro-automation?
Oh, definitely. I love it. DataRobot is awesome!
I love the fact that I can experiment with all of this data. I am a full-on pure adopter of any kind of automation that frees people to be creative (I speak as the owner of two robot vacuum cleaners and a robot lawnmower). I think it removes the boring, iterative parts of the job. Every single bit of data in the world is useful for a business purpose, and I think data scientists should spend their time innovating and creating new solutions to new business problems.
I love the challenge, and though ideas may fail, I now have the time to fail.