Tom Herold
Managing Director & Co-Founder at scalable minds
Berlin Metropolitan Area Germany
Summary
👤
Senior🎓
Top SchoolTom Herold is a managing director and co-founder with 14 years of experience building high-performance AI systems for large-scale biomedical image analysis. He leads Scalable Minds, shipping webKnossos for 3D dataset visualization and Voxelytics for automated biomedical image analysis, combining hands-on engineering with product leadership. Technically fluent across full-stack development and ML, he’s contributed frontend enhancements and evaluation improvements to open-source projects like Rasa and built React integrations for conversational UIs. Based in Berlin and trained at the Hasso Plattner Institute, he blends research-grade system design with practical deployment experience for life-sciences labs. An unassuming detail: he pairs deep image-reconstruction expertise with UX-minded frontend work, ensuring tools are both scientifically rigorous and usable.
14 years of coding experience
Master of Science (MSc), IT Sytems Engineering, Master of Science (MSc), IT Sytems Engineering at Hasso Plattner Institute
German, English
Stackoverflow
1,261reputation
78kreached
11answers
13questions
tensorflow
top-5%
Github Skills (24)
javascript10
python10
nlu10
machine-learning10
rasa10
natural-language-processing10
nlp10
react10
testing9
tensorflow9
chatbot9
html9
rest-api8
bot-framework8
sc7
Programming languages (11)
Github contributions (5)
scalableminds/chatroom
Mar 2018 - Jul 2021
React-based Chatroom Component for Rasa Stack
Role in this project:
Front-end Developer
Front-end DeveloperContributions:4 releases, 3 reviews, 40 commits in 3 years 4 months
Contributions summary:Tom primarily contributed to the front-end component of a React-based chatroom for the Rasa Stack. Their work involved adding new features, such as a welcome message and speech input integration, and refactoring the chatbot to utilize the REST channel for communication. They also addressed bug fixes, as seen with attachments and debug chat. The user demonstrated a good understanding of the chatroom's UI and its integration with the backend.
reactjsreactchatroomrasa
RasaHQ/rasa
Jan 2018 - Dec 2020
💬 Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants
Role in this project:ML Engineer
ML EngineerContributions:20 commits, 7 PRs, 15 comments in 2 years 11 months
Contributions summary:Tom primarily contributed to the improvement of the Rasa NLU evaluation reports. They implemented a confusion matrix and the "confused_with" attribute for response selection evaluation, indicating a focus on improving the interpretability of model performance. Further contributions included fixes to tests related to NLU and response selection, and added support for the markdown format with asterisks in training files.
nlupythonbotspeech-recognitionbotkit
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.