Expectations from Participants:

1. Please prepare a short-written overview covering:
   a. The health regulation context in your country: what is being regulated, by whom, and under what legal or institutional framework.
   b. The role and mandate of your organisation specifically: what powers does it have, what decisions does it make, and who does it report to.
   c. An honest estimate of AI literacy within your organisation: are there dedicated data or digital teams? Has AI been discussed at a strategic level? Are staff generally comfortable with data-driven approaches or is this relatively new territory? These short overviews will be shared with all participants in advance of the training day, so please write them with 17 1 The final Programme will be shared with participants that peer audience in mind. Reading them beforehand will help everyone arrive with a sense of the diversity of contexts in the room and make the group discussions significantly more productive.
2. Bring a laptop with Studio LLM installed
   An installation manual and step-by-step guide will be sent to all participants by email in advance of the training day. Please ensure Studio LLM is installed and working on your laptop before you arrive. If you encounter any difficulties during installation, please reach out to us so we can resolve this beforehand rather than on the day itself.
3. Prepare a 2-page dataset description
   Each participant must prepare a 2-page description of one dataset from their organisation that they believe has the potential for extended use or AI application. Please note that a “dataset” should be interpreted broadly. It does not need to be a structured database of quantitative data. Relevant examples include quality reports submitted by healthcare providers, inspection findings compiled over time, complaint logs, audit narratives, or accreditation documentation. If your organisation systematically collects any form of information, even partially in narrative or paper-based form, that is worth describing.

The dataset description can include the following information:

a. Basic information:
     a. Dataset name
     b. Who collects it?
     c. Why was it originally collected?
     d. Frequency of collection
     e. Format (structured, unstructured, mixed)
     f. Digital or partially paper-based?

b. Content description:
     g. Main variables (or types of information)
     h. Level of granularity (individual/entity/system level)
     i. Time coverage
     j. Estimated size

c. Data quality and constraints:
     k. Known limitations
     l. Missing data issues?
     m. Potential biases
     n. Legal/privacy constraints
     o. Access restrictions

d. Current use:
     p. How is it currently used?
     q. Who uses it?
     r. For what decisions?