REFRESH HISTORY

THE PROJECT AIM

This project aims to save development time for other ongoing projects or incidents within a business intelligence department at a reputable company, developers will be able to refer to a single source of information whilst being able to highlight unusual patterns and trends to identify blockers quicker.

The dashboard also consists of Linear Regression measures to predict refresh failures over the next couple of weeks to encourage tailored discussions on how to reduce failed dashboard refreshes or improve the fluidity of dataflows.

METHODOLOGY

Refresh history data was available on Power BI Web, however this needed to be extracted in JSON format to be able to process this data. I used Microsoft PowerShell to call Power Bi API functions to web-scrape the refresh history details online.

The department made use of OneDrive folders to keep track of version history, and workspaces for reports and dashboards, therefore the JSON data pulled was stored in an excel file to convert the data into a tabular format, and the file was scripted to upload to OneDrive, ensuring that whenever the end user runs the dashboard, the new file will overwrite the details contained in the initial file - this is a necessary feature due to the learned process of dealing with refresh failures, it is vital that developers know the current or live state of refreshes.

Power Automate Desktop was used to ensure the PowerShell script can be executed automatically without the end user’s interaction.