Presentation of the meeting (PDF)
Recording of the event
The programme
Supporting Research in Azure: An Insight into the Funded Projects
Presentation 1: Funding for Research with Azure Services
Speaker: Fabian Jusufi
To support research activities in Azure, the ZID funded projects in 2024 and 2025 with a total funding pot of € 20,000.00. In 2025, the ZID was able to fund 9 projects. These started at the beginning of the year, and some will continue into next year. The projects not only come from different scientific disciplines, but they also differ greatly in their approach, but almost all of them use the AI services in Azure: OpenAI Service and Azure AI Foundry. Due to the positive feedback from previous participants, the ZID is once again offering financial funding for 2026.
Presentation 2: Project-SciTextMatch: AI-Assisted Identification of Cross-University Collaboration Potential Aligned with Grant Calls
Speaker: Lilian Nowak
Project-SciTextMatch identifies collaboration potential via two complementary pipelines: a document-based matching pipeline (TF-IDF + SPECTER2) on publication titles/abstracts, which is validated by OpenAI’s GPT-4.1 and aligned with relevant grant calls; and a candidate-based pipeline using OpenAI deep-research agents across full PI lists. Outputs are integrated: overlaps are auto-accepted, and discrepancies are flagged for human review.
Presentation 3: Clarifying Without Bias: Chatbots in Experimental Economics
Speaker: Can Çelebi
This project develops large language model powered chatbots to support participants in online economic experiments. The goal is to improve subject understanding without introducing bias, by providing clarification and answering questions in the same neutral and procedural manner expected from a research assistant in a lab setting. The work explores how to build and refine such agents, including both single-agent setups and multi-agent workflows with role-specific prompts.
Jupyter Notebook
Presentation: Empowering Research with Jupyter Notebook
Speakers: Raman Ganguly, Gerald Teschl
Infrastructure Services for Research
Presentation: Project University Data Network 2025
Speaker: Ulrich Kiermayr