Leading Characteristics Contributing to Pupil Absence in England
With this project I wanted to utilise typical real-world data that is published in a non-ideal format. This was test of incorporating data from multiple 'single aspect' reports, many of which requiring conforming as they had evolved over time.
Additionally, visualising this data has led me to experimenting with the shape-map within Power BI and discovering that a relatively simple tool such as mapshaper had behemoths like QGIS beaten in the 'it just works' department.
I was able to experiment with passing a table of data, formatted as a text string within a DAX measure, to the Power Automate flow for the LLM to analyse. I was also able to use DAX to create a string of 'currently visible filter items' to create a context for matching, which would later be used to filter which analysis should be shown.
One aspect which will not be usable within this embedded report is the ability to trigger the Power Automate flow as the ability to share the flow relies upon the user's assignment to 'Users and Groups' within an organisation.
Whilst I will prepopulate a portion of the analysis, this method does work to allow internal users to analyse any possible visible combination of data and trigger new analysis when additional data is presented as part of a refresh.
You may note that visually, this report takes inspiration from the Power BI Core Visuals Vision Board, a great demonstration of project and task tracking.
The Department for Education produces additional data, such as staffing levels, which could be integrated at a later date to ascertain impact on pupil absence.

