Miriam Kuemmel
Co-Founder at Blue Guardrails
Berlin, Germany
Summary
🤩
Rockstar🎓
Top SchoolMiriam Kuemmel is a technical leader and co-founder focused on building production-ready LLM applications and tooling that make complex AI reliable and accessible. With seven years of experience bridging computational linguistics and applied language modeling, she helped deploy semantic search, RAG systems, and chatbots across industries at deepset and now leads Blue Guardrails to detect and prevent hallucinations in AI systems. She combines hands-on engineering—contributing to the popular Haystack framework by improving retrieval components and pipelines—with team leadership and direct customer engagement to turn research into practical solutions. Based in Berlin, Miriam is equally comfortable debugging why a RAG system fails and managing timelines to deliver working AI in production.
7 years of coding experience
6 years of employment as a software developer
Master of Arts - MA Speech and Language Processing, Master of Arts - MA Speech and Language Processing at University of KonstanzBachelor of Arts - BA German Literature and Linguistics, Bachelor of Arts - BA German Literature and Linguistics at FAU Erlangen-NürnbergResearch visitor, Research visitor at University of Massachusetts Amherst
Github Skills (8)
transformers10
retrieval-augmented-generation10
nlp10
large-language-models10
python10
question-answering10
machine-learning9
pytorch8
Programming languages (3)
Github contributions (5)
deepset-ai/haystack
Feb 2022 - Mar 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:
ML Engineer
ML EngineerContributions:13 reviews, 5 commits, 9 PRs in 1 month
Contributions summary:Miriam primarily focused on updating and refactoring code related to the integration and usage of different retrieval methods within the Haystack framework. Their contributions involved replacing the Dense Passage Retriever (DPR) with the EmbeddingRetriever, modifying tutorial notebooks to reflect these changes, and ensuring the correct functionality of pipelines. They also addressed issues related to the visualization of pipeline graphs and corrected minor code errors.
pythonbertfinetuningproduction-readyhuggingface-transformers
mkkuemmel/EllipsisDetection
Apr 2019 - Jun 2019
Contributions:14 pushes, 1 branch in 2 months
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.