Hannes Hapke is a seasoned machine learning engineer and author based in Portland, Oregon, with over a decade of experience delivering production-ready ML and NLP solutions. As a Principal Machine Learning Engineer at Digits since 2020, he leads scalable pipelines and model deployment across teams, drawing on a track record that spans SAP Concur, Cambia Health Solutions, and startup leadership. An active community builder, he co-authored O'Reilly's Building Machine Learning Pipelines and NLP in Action, and serves as a Google Developer Expert in Machine Learning and a Google Developer Advisory Board Member. His open-source contributions span Django back-end utilities, ML pipelines, and TensorFlow/TFX-based workflows, reflecting hands-on work from data ingestion to deployment. He has led cross-disciplinary teams and advised on AI strategy while maintaining hands-on coding with Python, Keras, and TensorFlow. In addition to corporate roles, he has led and advised startups like Wunderbar.ai and co-founded renooble, underpinning a blend of technical depth and entrepreneurial drive.
Code repository for the O'Reilly publication "Building Machine Learning Pipelines" by Hannes Hapke & Catherine Nelson
Role in this project:
Data Scientist
Contributions:2 releases, 3 reviews, 110 commits in 1 year 10 months
Contributions summary:Hannes contributed to the development of a machine learning pipeline, as evidenced by the addition of utility scripts for data splitting, and the creation of a Keras-based model experiment notebook. The primary focus of the work appears to be around data ingestion, preprocessing and model building. The user also focused on visualizing the model.
Contributions:19 commits, 5 PRs, 10 comments in 8 months
Contributions summary:Hannes primarily contributed to a TFX pipeline designed for processing and training sentiment analysis models using BERT. Their work involved updating the notebook to include the setup of an ALBERT model and integrating it with the existing pipeline. The user also focused on correcting and refining the model architecture and data preparation steps, including handling the input data structure for the ALBERT model. Furthermore, they adjusted pipeline configurations and package installations to ensure compatibility with the updated TF and TFX versions.
javaevents
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.