heidekrueger Goto Github PK

I'm passionate about using math, algorithms and data to enable smarter decision making, whether in business, policy or our private lives. Currently, I'm a Computer Science PhD student at TU Munich, doing research at the intersection of Artificial Intelligence, Algorithmic Game Theory and Microeconomics. In my research, I study applications of Multiagent Reinforcement Learning methods to compute nontrivial market equilibria, especially in auctions, and the mathematical underpinnings this entails. Previously, I worked as a Data Scientist in the telecomunications industry after earning my BSc and MSc degrees in applied Mathematics.

My expected graduation date for my PhD program is Fall/Winter 2022. After that I will be looking for Research Scientist / Research Engineer / AI Engineer / ML Engineer or (product-focused!) Data Scientist positions in industry. Ownership is very important to me, so I am currently not interested in 'Data Science Consultant' roles. I'm currently located in Munich, Germany. Primarily, I'm looking for positions that are on-site in Munich or remote/hybrid (European time zones), but I'm willing to relocate within Europe for the right opportunity.

You can follow me on Twitter, connect with me on LinkedIn or check out my academic profile at TUM.

In my free time, I enjoy making music (I'm a drummer in a band and play some guitar on my own), and exploring the beer gardens, lakes (via SUP) and mountains (via Snowboard) around Munich.

I'm passionate about (and have several years of professional experience in) the following frameworks:

• pytorch for deep learning -- and GPU-accelerated scientific computing in general.
• The tidyverse in R (e.g. dplyr, ggplot2, purrr, ...) for data analysis, vizualisation and statistical programming, as well as mlr3 and tidymodels for meta-ml in R.
• python for general software development. (I'm well versed in the Scipy stack for Data Science/ML (numpy, scipy, scikit-learn, matplotlib, plotnine,...), and have also worked on a few web apps in Django.

In the past, I've also worked on deep learning with TensorFlow and keras, Big Data with spark and scala, built Dashboards and data-centric WebApps using Tableau or RShiny, and written general-purpose code in Java, C# and C++, as well as some simulations with MATLAB. I know my way around Unix systems and git, I strive to write clean, performant, maintainable and we'll-documented code, and to follow DevOps best practices. I'm familiar with GitHub/ GitLab workflows and CI tools (especially GitLab CI). I have also worked extensively with RDBMSs (MySQL, Postgres, MS SQL, Oracle) in large DWH environments, and the cloud (mostly on AWS (EC2, S3, R53, AWS-CLI, SES, SageMaker, Lambda), but also some Azure).

Some technlogies I would love to check out in more detail, but haven't yet had a chance to work with extensively: JAX for differential programming; the julia project for statistical computing.

A good start would be bnelearn, a python library for equilibrium computation in Bayesian games that I originally developed during my PhD, and that I now maintain together with multiple collaborators. The library contains highly performant parallel implementations of many markets studied in the Auction Theory literarture, and it has enabled research that has appeared in journals such as Nature Machine Intelligence, the INFORMS Journal on Computing, and AI conferences like AAMAS and AAAI. You can find the latest public release at https://github.com/heidekrueger/bnelearn.

I also encourage you to also check out my solutions to Google's semi-secret foo-bar recruiting challenge, that I finished recently. (Levels 4.1 and 4.2 are a good place to start.)