Timo Möller
Co-Founder at deepset
Berlin, Germany
Summary
🤩
Rockstar🎓
Top SchoolTimo Möller is a co-founder and NLP engineer with 13 years of experience building production-ready ML systems and open-source frameworks from Berlin. He co-created deepset’s FARM and Haystack projects—tools widely used for question answering, retrieval-augmented generation, and neural search—and has driven key improvements around QA inference, multi-GPU support, and robust “no answer” handling. His background spans applied ML roles at plista and freelance projects for Springer Nature and others, delivering measurable business impact like ad revenue uplift and improved recommendation CTR. Passionate about open source and ethical AI, he focuses on bringing research-grade NLP into industry deployments and has a computational neuroscience MSc that informs his principled approach to model design.
13 years of coding experience
2 years of employment as a software developer
Bachelor of Science - BS, Data Science and Knowledge Engineering, Bachelor of Science - BS, Data Science and Knowledge Engineering at Universiteit MaastrichtMaster of Science (M.Sc.), computational neuroscience, Master of Science (M.Sc.), computational neuroscience at Technische Universität Berlin
Github Skills (12)
transformers10
pytorch10
machine-learning10
transfer-learning10
retrieval-augmented-generation10
nlp10
large-language-models10
python10
question-answering10
bert9
ml9
deep-learning9
Programming languages (9)
Github contributions (5)
deepset-ai/FARM
Jul 2019 - Sep 2021
:house_with_garden: Fast & easy transfer learning for NLP. Harvesting language models for the industry. Focus on Question Answering.
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:10 releases, 117 reviews, 269 commits in 2 years 1 month
Contributions summary:Timo made several code changes focused on question answering (QA) functionality within the FARM framework. Their contributions included bug fixes related to QA inference mode, adjustments to layer dimensions and answer validation, as well as returning the correct answer from extracted text. They also added support for multi-GPU usage in experiments and examples. Furthermore, they worked on refactoring QA code for improved efficiency and accuracy.
bertxlnet-pytorchnatural-language-processingtransformersharvesting
deepset-ai/haystack
Dec 2019 - Oct 2022
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:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:92 reviews, 75 commits, 99 PRs in 2 years 10 months
Contributions summary:Timo's commits primarily focus on enhancing the `FARMReader` component within the Haystack framework. They implemented and refined features related to "no answer" handling and confidence scoring for question-answering tasks. The contributions involve modifying the core logic of the `FARMReader` to include metrics for the "no answer" possibility and incorporate semantic answer similarity. They also addressed code improvements and typo fixes within the tutorials and documentation.
pythonbertfinetuningproduction-readyhuggingface-transformers
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.