Research Data Workflows
and LLMs

Session 1 · Making Estate Materials Digitally Accessible
Research Data Workflows in Stefan Zweig Digital
SlidesPDFLecture notesPDFLecture notes written with AI assistance.
Learning objectives
- Understand what research data is and why it matters.
- Understand research data workflows and the role of metadata standards.
- Understand the principles of the Semantic Web and Linked Open Data.
Hands-on materials
All materials · ZIP · 2.5 MBDescribe the “Schulnachricht” in XML
From XML to IIIF with Python
Encode the Schulnachricht in TEI XML with LLM Assistance
Represent a Source Image in RDF with LLM Assistance

Session 2 · Large Language Models for Research Data Workflows
An Introduction
SlidesPDFLecture notesPDFLecture notes written with AI assistance. Work in progress: sections may be incomplete and are being revised.
Learning objectives
- Develop a basic understanding of large language models (LLMs).
- Understand the fundamentals of prompt and context engineering.
- Gain initial hands-on experience using LLMs in research data workflows.
Hands-on materials
All materials · ZIP · 52.0 MBTranscribe a Facsimile with Gemini 3.8 Flash
Exploring Research Data with LLMs
From Facsimiles to TEI XML

Sessions 3 and 4 · Large Language Models for Research Data Workflows
Hands-On Practice
SlidesPDFLecture notesPDFLecture notes written with AI assistance. Work in progress: sections may be incomplete and are being revised.
Learning objectives
- Use LLMs for coding and build small research tools through Promptotyping.
- Understand AI agents and the role of an AI harness.
- Apply context and knowledge engineering to research tasks.
- Practise agentic engineering by inspecting tool use, checking results and refining instructions.
Hands-on materials
All materials · ZIP · 4.8 MBAI Harness and M³GIM Data
Promptotyping Project: A Small Digital Edition or Research Dashboard
Tools
From XML to IIIF viewer
Built for the Session 1 hands-on with Claude Fable 5.1. Select your XML description or generated IIIF manifest with its page images and see them as a digital object in Mirador, entirely in your browser.