Lunch Bites 2026: KadiAssistant – An AI Assistant for Research Data Management
presented by Adrian Cierpka (KIT)
How can researchers efficiently find and connect knowledge across heterogeneous and privacy-sensitive research data? We present KadiAssistant, an AI assistant integrated into the Kadi research data ecosystem. By combining a self-hosted large language model with privacy-preserving semantic search, KadiAssistant can access and synthesize information from diverse research data while strictly respecting fine-grained access permissions. The system bridges disciplinary terminology and standards, lowers barriers to data discovery, and supports the Findable pillar of FAIR data principles.
Program
This Lunch Bite introduces KadiAssistant, an AI-powered assistant designed to support Research Data Management throughout the research lifecycle.
Registration
You can register in the form below.
Further Information
- Paper preprint: https://arxiv.org/abs/2605.18850
- Authors: Adrian Cierpka (ORCID), Mohammad Shafiqul Islam (ORCID), Johannes Steinhülb (ORCID), Eric Dietriche Sesso Domtchoueng (ORCID), Michael Selzer (ORCID), Arnd Koeppe (ORCID)
- Related project:https://www.postlithiumstorage.org/de/
About NFDI4ING Lunch Bites
NFDI4ING Lunch Bites is a monthly online series featuring short presentations and discussions on Research Data Management in engineering. Each session consists of a 15-minute talk and a 15-minute Q&A, offering practical insights into research projects, workflows, tools, and services from the engineering community. The series takes place on the first Wednesday of every month.
More information on our Lunch Bites can be found here.
Registration form