Karl Richter
Founding Engineer at Langdock (YC S23)
Munich, Bavaria, Germany
Summary
🤩
Rockstar🎓
Top SchoolKarl Richter is a Founding Engineer based in Munich with 8 years of experience building data-driven systems and startups. He has deep hands-on expertise in large-scale data engineering and PySpark pipelines from multi-year work at Mercedes‑Benz, where he progressed to leading a team that delivered a collaborative-filtering recommender. Karl blends product-focused engineering with operational excellence—evident from his contributions to open-source forecasting tooling (neural_prophet) where he improved testability and modernized core data handling. As a founder at Langdock (YC S23) and earlier co-founder of WILM&CO, he repeatedly ships end-to-end platforms for fast-moving teams. He pairs formal training from TUM and research time at Stanford with practical R&D experience in NLP and voice assistants, giving him a strong footing across ML, backend, and DevOps. Known for pragmatic refactors and performance-minded code changes, he favors durable, testable systems over clever one-offs.
8 years of coding experience
5 years of employment as a software developer
Honours degree, Technology Management, Honours degree, Technology Management at Center for Digital Technology and Management (CDTM)M. Sc., Information Systems, M. Sc., Information Systems at Technical University of MunichResearch, Research at Stanford University
German, English, French
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (11)
pandas10
refactoring10
pytest10
python10
testing10
time-series9
deep-learning8
forecast8
deeplearning-ai8
forecasting8
machine-learning7
Programming languages (4)
Github contributions (5)
ourownstory/neural_prophet
Jun 2022 - Jan 2023
NeuralProphet: A simple forecasting package
Role in this project:
Backend & DevOps Engineer
Backend & DevOps EngineerContributions:101 reviews, 242 commits, 100 PRs in 7 months
Contributions summary:Karl focused on addressing pytest warnings within the codebase by refactoring the code to replace deprecated functions with modern, efficient alternatives. They updated the code in multiple files to use `pd.concat`, `frame.loc`, and `pd.assign`, improving code clarity and performance. They also made changes to the benchmarking and forecaster modules. This work suggests a focus on maintaining the code's quality, efficiency, and testability.
forecastingneuralprophetpythonforecasttime-series
karl-richter/neural_prophet
Jun 2022 - May 2023
NeuralProphet: A simple forecasting package
Contributions:6 PRs, 135 pushes, 14 branches in 11 months
pytorchforecastingneuralprophetpythondeep-learning
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.