Petra Axolotl

Interview with Petra Axoloti – November 2017

Math Central (MC): Thank you Petra for agreeing to take the time for this interview. Start by telling us your job title and your employer?

PETRA : I am a Data Scientist for Platterz.

MC: Can you describe your job?

PETRA: Platterz is a tech startup that makes corporate catering easier by assembling the right meals from the right restaurants, given constraints such as number of people to feed, budget, and dietary restrictions. Ideally, we do not want office managers to spend any time thinking what to order to feed their employees. It should be done by the Machine.

Mathematically the problem is NP-complete. Therefore, we try to find good meals, instead of the best meals, by following certain heuristics and breaking down the problem into several more tractable sub-problems.

I am one of the two data scientists designing and implementing the algorithm so that no human intervention is needed to assemble meals. The other data scientist (Lev Naiman) has a PhD in Computer Science from University of Toronto.

MC: Tell us a little about your background and education.

PETRA: I moved from one place to another as a child but consider Leiden, The Netherlands my home town. I have a Master’s degree in Computer Science from Leiden University (Netherlands) and an MBA from the Wharton School (Pennsylvania).

I had had several jobs before joining Platterz as a Data Scientist.

Management Consultant at A.T. Kearney, Amsterdam, Netherlands
Quantitative Researcher at Robeco, Rotterdam, Netherlands
Equity Analyst at Robeco, Rotterdam, Netherlands
Management Consultant at ING, Amsterdam, Netherlands
Data Scientist at Twitter, San Francisco, California
I am also a mentor for a Coursera Course Introduction into General Theory of Relativity.

MC: You mention a degree in Computer Science and an MBA but you also have a strong interest in Physics. Can you say something about your interest in Physics?

PETRA: Physics answers the most fundamental questions about the reality. Of course you can be completely sceptical like Descartes about reality. But physics is about as far as we can go if we assume the external world is real and understandable (i.e. everything is governed by deterministic or at least probabilistic laws).

To me personally, physics strikes the right balance between complete abstraction of certain fields of mathematics and applicability of engineering. Physics allows me to visual mathematics.

MC: When did you decide to make a career in an area involving so much mathematics?

PETRA: I enjoyed abstract thinking. As a quantitative researcher, I developed pricing models for credit default swaps, interest rate derivatives, etc., using techniques such as stochastic calculus and Markov simulations. I used statistical methods to come up with new trading strategies.

The easiest, and in fact somewhat boring, part of my work as a Data Scientist is the so-called Machine Learning, be it simple logistic regressions or more sophisticated multi-layer neural networks. What really interests me is to solve problems for which there are no existing solutions and which involves creative mathematical modeling. One such example is how to predict customer retention by modeling each customer as a coin-tossing machine with his/her hidden probability of staying a customer (let’s call it theta), and the whole cohort of customers as a distribution of theta’s described by a Gamma distribution.

MC: What do you do when you are not working? What are your hobbies and other interests?

PETRA: I spend a lot of time in the weekend working on problems in General Relativity and Quantum Physics. My personal website (www.petraaxolotl.com) is currently 100% dedicated to physics and mathematics relevant to it.

My other passion is hiking, to be far away from cities and be emerged in nature and close to wild animals and plants.

My daily routine also includes reading at least five lines of Latin, some Greek, and watching 20:00 o’clock evening news in French on France 2.