Thomas Stadelmann
Lead AI Engineer at deepset
Nuremberg, Bavaria, Germany
Summary
🤩
Rockstar🎓
Top SchoolThomas Stadelmann is a Lead AI Engineer based in Nuremberg with six years of experience building production-ready NLP and ML systems, currently leading AI initiatives at deepset. He moved up through roles from Open Source Engineer to lead, contributing substantial code and tutorials to the widely used Haystack framework—particularly around document classification and indexing pipelines—to make retrieval-augmented QA and semantic search more robust and accessible. His background blends applied data science and software engineering from DATEV with an academic foundation in Information Systems (MS, BS), enabling him to bridge research-grade NLP techniques and pragmatic engineering. Notably, he has hands-on experience authoring and maintaining Jupyter notebooks for the "NLP with Transformers" book, demonstrating a knack for clear, reproducible examples that help others adopt complex QA architectures.
5 years of coding experience
11 years of employment as a software developer
Bachelor of Science - BS, Information Systems, Bachelor of Science - BS, Information Systems at Baden-Wuerttemberg Cooperative State University (DHBW)Master of Science - MS, Information Systems, Master of Science - MS, Information Systems at Otto-Friedrich-Universität Bamberg
Github Skills (15)
transformers10
pytorch10
machine-learning10
lm10
jupyter-notebook10
retrieval-augmented-generation10
nlp10
large-language-models10
haystack10
python10
question-answering10
llm10
data-science9
generative-ai8
chatgpt8
Programming languages (7)
Github contributions (5)
deepset-ai/haystack
Nov 2021 - Jan 2023
AI orchestration framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your data. With advanced retrieval methods, it's best suited for building RAG, question answering, semantic search or conversational agent chatbots.
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 586 reviews, 200 commits in 1 year 2 months
Contributions summary:Thomas contributed significantly to the development of a document classification system within the Haystack framework. Their work involved creating a tutorial for a `DocumentClassifier` at index time, enhancing the `TransformersDocumentClassifier`, adding notebook functionality and testing for batch size usage. They also improved the flexibility of the indexing pipeline and extended the tutorial with zero-shot classification and pipeline integration.
pythonbertfinetuningproduction-readyhuggingface-transformers
nlp-with-transformers/notebooks
May 2022 - Jun 2022
Jupyter notebooks for the Natural Language Processing with Transformers book
Role in this project:Data Scientist
Data ScientistContributions:5 commits, 1 PR, 6 comments in 1 month
Contributions summary:Thomas primarily contributed to the Jupyter notebooks associated with the "Natural Language Processing with Transformers" book, specifically focusing on Question Answering (QA) tasks. Their contributions involved adding and updating code related to different QA system implementations, likely incorporating various techniques like span classification and the retriever-reader architecture. The commits show updates to incorporate feedback and adapt the notebooks to Haystack library changes. These changes suggest a focus on applying and explaining NLP concepts within a practical QA framework.
nlptransformerspythonlanguage-processingnatural-language
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.