Connect with us

Pre-Conference Workshop FD3

FD3 - Multi‑LLM Hackathon for Strategic Intelligence in Education & Training

Date Wednesday, 2 Dec Time   –    Price: 195.00 € Status: places available

OEB speaker Paul Bacsich

Paul Bacsich

Managing Director and Chief Researcher, Matic Media Ltd

OEB speaker Margaret Korosec

Margaret Korosec

Director of Digital Education and Learning Innovation, University of Leeds

OEB speaker Brian Mulligan

Brian Mulligan

Consultant, Universal Learning Systems

How can multiple Large Language Models be used to produce credible, decision-useful research in a limited timeframe? This hands-on workshop is structured as an AI Evidence Sprint, not a coding event, where participants work in small groups to investigate a strategic question relevant to education and training and produce a structured, referenced and verified analytical briefing.

This is a Bring Your Own LLM workshop: working with their own LLM toolkit, participants will define a research scope, establish inclusion and exclusion criteria, develop a structured outline and use AI to collect, analyse and synthesise relevant sources. The workshop focuses not only on generating content, but on developing reliable research workflows across different LLMs and tools. Participants will compare strengths and weaknesses of different approaches and learn how to identify and address uncertainty, hallucinations and weak evidence.

A mandatory verification process will be used throughout the sprint, including cross-checking at least five key claims through a second LLM, source checking or manual verification. Each group will work towards producing an analytical briefing of at least 10 pages, with a one-page executive summary, referenced claims and a methods and limitations section. Participants will leave with practical experience in AI-assisted research and transferable workflows for producing evidence-based strategic intelligence in their own organisations.

Agenda:

  • Welcome & Introduction to the AI Evidence Sprint (10 min)
  • Recent Examples of AI-Assisted Research & Reporting (20 min)
  • Group Formation & Project Selection (30 min)
  • Markdown-Based Workflows for Report Generation (45 min)
  • Q&A (15 min)
  • Coffee Break & Networking (30 min)
  • Refining Research Scope & Initial Testing (30 min)
  • Lunch Break (120 min)
  • Hands-on AI Evidence Sprint: Research, Analysis & Verification (120 min)
  • Team Elevator Pitches & Sharing (45 min)
  • Final Reflections & Closing (15 min)

Target Audience: Researchers, Education and Training Professionals, Higher Education Leaders, Policy and Strategy Professionals, Learning and Innovation Professionals, Educational Technology Professionals

Target Audience Sector: Secondary Education, TVET / Vocational Education, Higher Education, Workplace Learning, Public Sector

Prerequisite Knowledge: Advanced experience with at least one leading SaaS LLM toolkit is required. Participants must have an active commercial subscription to at least one frontier LLM capable of working with Word documents and Markdown, and should be comfortable using AI tools for research and document creation. Participants must bring a Windows or Mac laptop with Microsoft Word. No programming experience is required. Participants must not use confidential or access-restricted data during the workshop and are expected to comply with the conference’s GDPR and AI guidelines.

This is a Bring Your Own LLM workshop. Facilitators will support research methodology and workflow design, not detailed software troubleshooting.

Expected Outcomes:

  • Apply a structured workflow for AI-assisted research, from defining scope and research criteria through to synthesis and reporting.
  • Explore effective cross-LLM and Markdown-based workflows for creating, reviewing, updating and extending structured research documents.
  • Apply good practice for AI-assisted research, including source verification, hallucination reduction and the clear documentation of uncertain or contested evidence.
  • Evaluate the strengths and limitations of different LLMs and supporting tools, including document, citation and research workflows
  • Explore approaches to document import, multilingual research and structured data collection.
  • Produce a structured analytical briefing of at least 10 pages, including referenced claims, a one-page executive summary, and documented methods and limitations.