At Deloitte, I initially followed the traditional audit steps to analyze accounts and present audit findings until HQ introduced a new analytical platform, Power BI, to better visualize our findings.
Selected as the first batch to receive Power BI training, I realized that Power BI couldn’t be used unless the template that the HQ provided was fully completed, and unfortunately, many companies lacked the necessary sufficient/unified data. To solve this, I learned exactly what data is needed to compete every column on the template and to explain it to the client, and once I figured out the needed accounts and transaction details, I had to deeply understand their whole operational flow and ERP system to request the right data to match the template. This usually required several iterations with different departments, since each department could only provide data from the ERP system that they were familiar with. Eventually, after multiple failed attempts, I finally gathered unified data.
In the end, the presentation was well-received by the client and my partner. The visualization that I created even became a template for other Deloitte audit teams. While I was glad to make this contribution, I recognized that such templates designed for auditors that do not have programming ability would have limited effectiveness. If I had Python skills in the first place, I could have adjusted the setup in Power BI to create visualizations without having unified data, and I am eager to start building those skills at Columbia.