Skip to content
View heidekrueger's full-sized avatar
18 followers · 20 following

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Starstruck

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Starstruck

Block or report heidekrueger

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
heidekrueger/README.md

Hi, I'm Stefan 👋

I'm a product-minded AI engineer based in Munich, Germany. I'm passionate about using ML, math, algorithms, and data to enable smarter decision-making, whether in business, policy, or our private lives.

I'm currently working as a Senior AI Research Engineer at Buynomics. Previously, I led the development of an internal GenAI LegalTech application at Munich Re, and worked as a Data Scientist at Telefónica Germany. I have also completed some freelancing projects around AI, ML, and agentic systems.

I hold a PhD in Computer Science from TU Munich. My research focused on the intersection of Artificial Intelligence, Algorithmic Game Theory, and Microeconomics. In particular, I studied applications of deep multiagent reinforcement learning methods to compute nontrivial market equilibria, especially in auctions, and the mathematical underpinnings of learning dynamics in such markets.

You can connect with me on Twitter or LinkedIn.

In my free time, I enjoy spending time with my family, playing rock music (drums + some guitar), snowboarding and biking, or vibe-coding tools for personal use.

If you're here as a potential employer who wants to see some of my (pre-LLM era) code:

Most of my recent software projects are proprietary, unreleased, or private by design. Sorry. :( I find that the hobby software projects I'm drawn to (e.g. for parenting or personal finance) tend to involve data or information that I can't or don't want to (re)distribute. Reach out to me individually if you want to know more.

If you're willing to go further back (to a time when code was still written by hand!), a good start would be bnelearn, a Python library for numerical equilibrium computation in Bayesian games that I originally developed during my PhD, and then maintained with multiple collaborators. The library contains highly performant parallel implementations of many markets studied in the Auction Theory literarture, and it has enabled research (by me, as well as others) that has appeared in journals such as Nature Machine Intelligence, the INFORMS Journal on Computing, Management Science, and AI conferences like ICML, AAMAS, and AAAI. You can find the latest public release at https://github.com/heidekrueger/bnelearn.

I encourage you to also check out my solutions to Google's semi-secret foobar recruiting challenge, which I completed sometime in 2022. (Levels 4.1 and 4.2 are a good place to start.)

Pinned Loading

  1. bnelearn bnelearn Public

    A Framework for Equilibrium Learning in Sealed-Bid Auctions

    Jupyter Notebook 24 2

  2. f00b4r-Challenge f00b4r-Challenge Public

    Python