Stefan Falk
Product Owner & Senior Software Engineer at Dynatrace
Klagenfurt, Carinthia, Austria
Summary
🤩
Rockstar🎓
Top SchoolStefan Falk is a Product Owner and Senior Software Engineer based in Klagenfurt, Austria, with seven years of professional experience building ML-driven products and production software. He transitioned from hands-on roles in machine learning and speech translation at iTranslate to product and engineering leadership at Dynatrace, bridging model development and product delivery. His contributions to the widely used tensorflow/tensor2tensor project show practical expertise in model lifecycle tooling, data handling, and speech recognition metrics like word error rate. Stefan combines a solid academic foundation in computer science from TU Graz with experience across data science, embedded software, and test tooling from Infineon and Know‑Center. Comfortable both coding and shaping roadmaps, he brings a pragmatic focus on shipping reliable ML systems that scale in enterprise contexts. A former NATO peacekeeper and student leader, he also offers disciplined teamwork and operational perspective beyond pure engineering.
7 years of coding experience
10 years of employment as a software developer
Master's degree, Computer Science, Master's degree, Computer Science at Technische Universität Graz
English, German
Stackoverflow
24,667reputation
3.5mreached
240answers
758questions
spring-boot
top-5%
python
top-5%
Github Skills (17)
python10
machine-learning10
deep-learning10
tensorflow10
speech-recognition10
data-pipelines9
spring-boot9
data-pipeline9
nlp8
spring-data-jpa6
angular6
pep6
spring6
java6
pandas6
Programming languages (10)
Github contributions (5)
tensorflow/tensor2tensor
Sep 2018 - Nov 2018
Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research.
Role in this project:
ML Engineer
ML EngineerContributions:9 commits, 7 PRs, 45 comments in 2 months
Contributions summary:Stefan primarily focused on enhancing the `tensor2tensor` library by modifying and extending its functionalities. Their contributions include adding features, such as a method to load models from registry, and incorporating hooks to pass hyperparameters to training and evaluation processes. Furthermore, they addressed issues with data handling, particularly regarding the passing of the data directory to feature encoders, and contributed to improving metrics, notably the implementation of word error rate for speech recognition tasks.
pytorchautoencoderdeep-learningmachine-translationreinforcement-learning
stefan-falk/tensor2tensor
Nov 2018 - Jun 2019
Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research.
Contributions:2 pushes in 6 months
pytorchdeep-learningaccessibletorchml
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.