Beschreibung

An der Georg-August-Universität Göttingen Stiftung Öffentlichen Rechts – Institut für Informatik sind ab dem 01.12.2025 2 Stellen alsPhD Researchers — LLM Pre-Training & RAG (all genders welcome)- Entgeltgruppe 13 TV-L - in Vollzeit (teilzeitgeeignet) zu besetzen. Die Stellen sind befristet bis 30.11.2028.The Chair of Scientific Information Analytics, headed by Prof. Dr. Bela Gipp (GippLab — https://gipplab.uni-goettingen.de), conducts research at the intersection of computer science, data science, and information science. Natural language processing (NLP) based on Large Language Models (LLM) is a particular research focus of the chair.

As part of the research project LibraryAI — LLM Pre-Training and RAG for Libraries (more information), ​we are looking for two full-time positions for PhD researchers (all genders welcome).

Short Project Description: Libraries have long evolved—from scrolls to books to computerized catalogs and web-scale discovery services. With generative AI, the next transformation is already taking shape. The LibraryAI project proposes to design, develop, and test a virtual research assistant powered by Large Language Models and Retrieval-Augmented Generation (RAG) to support library users.
This project is a joint effort between the University of Göttingen (UGoe), the Göttingen State and University Library (SUB), and the GWDG (Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen). Our multidisciplinary team brings together expertise in library science, computer science, and data management. Moreover, the project can leverage significant data and high-performance computing resources. This blend of domain-specific knowledge, technical skill, and data infrastructure will empower us to devise, implement, and deploy a capable, robust, and efficient language model along with a user-friendly interface.
LibraryAI will be agnostic to the underlying library systems and collections, feature a modular architecture, and be made openly accessible to ensure its adaptability for other libraries.

Areas of Responsibility

Adapt and fine-tune existing open-source language models using transfer learning to meet the specific needs of the library and its diverse scientific collections.
Integrate the LLM with an information retrieval system that accesses the library's most current content, ensuring answers are up-to-date and factually grounded.
Design, implement, and test an integrated solution that combines LLMs, RAG, and agents to use external tools and APIs, enabling it to query the library catalogue, check for book availability, access digital repositories, and answer questions about library services.
Develop and train models based on PPO, GRPO, and RLVR to align models to human preferences.
Contribute to front- and back-end development for the project.
Collaborate with an interdisciplinary team of subject matter experts, data specialists, and project partners, including the GWDG.
Contribute to the development of a robust and sustainable MLOps pipeline for real-world use in libraries.
Engage in the development of project frameworks and support partners in deploying models within user-facing web applications.
Publish your research findings at leading NLP conferences (e.g., ACL, EMNLP, NAACL).
Conduct teaching in accordance with the requirements of the PhD program in computer science at the University of Göttingen (low workload).
Supervise bachelor's and master's students working on related projects.


Your Profile


A master's degree with a grade of very good (or equivalent) in computer science, computational linguistics, data science, or a related discipline.
Strong programming skills, particularly in Python.
Hands-on experience with deep learning frameworks (e.g., PyTorch) and NLP libraries (e.g., Hugging Face).
Solid foundation in NLP downstream tasks (e.g., NER, tokenization, word sense disambiguation).
Professional fluency (C1) in German and English, both written and spoken.
Communicative and teamwork-oriented mindset.


We Offer


A stimulating, interdisciplinary, and international research environment at one of Germany’s top-tier universities.
The chance to contribute to a high-impact project at the intersection of AI, linguistics, and cultural heritage.
Close, supportive supervision to foster both your academic and personal development toward a PhD.
Access to cutting-edge HPC resources, such as a GPU cluster with NVIDIA A100 GPUs, is essential for large-scale model training.
Funding to present your research at premier international conferences and engage with the global scientific community.
A full-time position with a competitive salary under the German public service scale (TV-L), with the option of part-time employment.
Office space and technical equipment. The position will be located in Göttingen on-site.

This role requires regular on-site presence in Göttingen; remote work is possible to a limited extent.Die Universität Göttingen strebt in den Bereichen, in denen Frauen unterrepräsentiert sind, eine Erhöhung des Frauenanteils an und fordert daher qualifizierte Frauen nachdrücklich zur Bewerbung auf. Sie versteht sich zudem als familienfreundliche Hochschule und fördert die Vereinbarkeit von Wissenschaft/Beruf und Familie. Der beruflichen Teilhabe von schwerbehinderten Beschäftigten sieht sich die Universität in besondere Weise verpflichtet und begrüßt deshalb Bewerbungen schwerbehinderter Menschen. Bei gleicher Qualifikation erhalten Bewerbungen von Menschen mit Schwerbehinderung den Vorzug. Eine Behinderung bzw. Gleichstellung ist zur Wahrung der Interessen bereits in die Bewerbung aufzunehmen.Bitte reichen Sie Ihre aussagekräftige Bewerbung mit allen wichtigen Unterlagen bis zum 29.10.2025 ausschließlich über das Bewerbungsportal http://obp.uni-goettingen.de/de-de/OBF/Index/76074 ein. Auskunft erteilt Herr PD Dr. Terry Lima Ruas, E-Mail: ruas@uni-goettingen.de, Tel. +49 551 39 25835Hinweis: Wir weisen darauf hin, dass die Einreichung der Bewerbung eine datenschutzrechtliche Einwilligung in die Verarbeitung Ihrer Bewerbungsdaten durch uns darstellt. Näheres zur Rechtsgrundlage und Datenverwendung finden Sie im Hinweisblatt zur Datenschutzgrundverordnung (DSGVO)