Lukas Geiger
Deep Learning Scientist at Plumerai
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolLukas Geiger is a Deep Learning Scientist based in London with 10 years of experience bridging astroparticle physics, ML research, and production-grade software engineering. Currently at Plumerai, he applies rigorous scientific thinking to efficient deep learning models and tooling, with a strong track record contributing to flagship open-source projects like TensorFlow Addons, Keras, SciPy and JAX. His contributions span performance optimizations, cross-platform build and CI work (including macOS M1 wheel support), and dataset engineering for TensorFlow Datasets, showing fluency from low-level numerical code to developer-facing libraries. Prior roles at nteract and extensive OSS mentoring reflect his focus on interactive computing and developer experience. Colleagues value him for turning subtle numerical and compatibility issues into robust, test-covered fixes—an approach shaped by his background in astroparticle physics.
10 years of coding experience
3 years of employment as a software developer
Bachelor of Science (B.Sc.), Physik, Bachelor of Science (B.Sc.), Physik at RWTH Aachen University
German, English
Github Skills (90)
atom-editor10
apim10
python10
image-processing10
testing10
windows10
language-server-protocol10
datasets10
deep-learning10
build-automation10
linter10
javascript10
static-analysis10
type-check10
data-model10
Programming languages (23)
Github contributions (5)
nteract/hydrogen
Feb 2016 - Nov 2018
:atom: Run code interactively, inspect data, and plot. All the power of Jupyter kernels, inside your favorite text editor.
Role in this project:
Full-stack Developer
Full-stack DeveloperContributions:47 releases, 651 commits, 486 PRs in 2 years 9 months
Contributions summary:Lukas primarily focused on adding features and enhancing the functionality of the hydrogen project, which seems to be an interactive coding and data inspection tool built into a text editor. Their contributions include implementing the ability to run a whole file with a single command and adding support for running code cells, along with the modification of menus and keybindings. Furthermore, they incorporated improvements like styling the watch sidebar and result bubbles to improve user interface and experience.
nteractpythonjupyterlab-extensionjupyter-notebookinspect
zeromq/zeromq.js
Sep 2016 - Nov 2019
:zap: Node.js bindings to the ØMQ library
Role in this project:Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:24 releases, 150 commits, 113 PRs in 3 years 2 months
Contributions summary:Lukas primarily contributed to the build and deployment processes for the zeromq.js project. They addressed build warnings, fixed issues related to dependencies, and synced the project with upstream changes. Furthermore, the user was involved in updating the Windows build configuration, including static linking and incorporating the latest VS toolset, and setting up continuous integration to deploy documentation. They also refactored build scripts and upgraded the project's libzmq version.
node-jszapnodejsjs-bindings
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.