Introduction to AI-assisted workflows for unstructured data: Demonstrating programmatic use of LLMs

Research projects collect and generate unstructured data, from interview transcripts and social media feeds to PDFs, audio, and image files. While manual coding of this data is essential, it presents significant scaling challenges. In this workshop, we introduce and demonstrate how Large Language Models (LLMs) can be used programmatically to facilitate and scale analysis. Our goal is to look beyond standard chatbot interfaces, demystify the underlying technology, and showcase the broader technical landscape of these tools. This is a demonstration-focused workshop rather than a hands-on coding session. While we will showcase programmatic implementations and technical workflows, participants will not be writing or executing code themselves.

For all researchers, staff and research students, and research support staff.

Mode of delivery: Online

Dates:

• Tuesday 3 March 2026
• Friday 14 August 2026

• Understand the core technical concepts of building an LLM-assisted research workflow and how these tools can be broadly applied across diverse research questions
• Observe programmatic implementations of LLMs in action using the Python programming language
• Identify key ethical risks, including bias, data privacy, and model hallucinations and sycophancy
• Critically evaluate and validate LLM outputs
• Determine whether these methods, as well as the associated skills, tools, and permissions, align with your research context

• Further workshops on this subject include Embedding Large Language Models (LLMs) into qualitative research workflows.
• Researchers and postdoctoral students are encouraged to engage with their peers by joining Hacky Hour to post questions and exchange ideas, and to contact Centre for eResearch (CeR) staff.

Research Data Support Services
Email: researchdata@auckland.ac.nz

Toby Johnson
eResearch Engagement Specialist, Centre for eResearch
Email: toby.johnson@auckland.ac.nz