POWER BI PROJECTS
In this section, you’ll find some of my first Power BI projects, which I started after gaining experience with Python and SQL. Each project was developed using different types of datasets from Kaggle.
These projects are a bit smaller, since most of them are built on a single table. However, many of these tables include a large number of columns, which allows for interesting analyses and exploration through DAX and Power Query.
Feel free to explore them and see how my skills have grown step by step. And don’t forget to check my main portfolio section, where I showcase my TOP projects—the ones that best highlight my current capabilities as a Data Analyst.
Title: Sales Performance — Superstore (2015–2018)
Context: Five-page Power BI dashboard analyzing sales, products, shipping, and geography for a U.S. retail dataset (Kaggle).
Learning / main challenge: My first solo Power BI project: full end-to-end build, custom page templates, a proper Date table, order-level vs line-level logic (ALLEXCEPT), and a SUMMARIZE helper table to enable map legends by sales category.
View full project by clicking –>
Title: Airline Customer Satisfaction.
Context: Five-page Power BI report on 129,880 survey responses (22 fields) exploring satisfaction across in-flight vs pre/post-flight services, delays, distance, class, and demographics (Kaggle).
Learning / main challenge: Cleaning quirks, deciding that 0 = missing, reshaping 14 satisfaction fields with unpivot, and building robust DAX (% Survey Completion, REMOVEFILTERS vs ALL) + Python correlation heatmap.
<– View full project by clicking
Title: IBM HR Analytics — Employee Attrition
Context: Power BI report exploring headcount, pay, satisfaction (1–4 scales), and attrition drivers using IBM’s fictional HR dataset (Kaggle).
Learning / main challenge: Advanced bookmark system (including a filter pop-up that preserves slicers), a clean design system (palette + sidebar + icons), and robust % measures using
REMOVEFILTERSvsALL/ALLEXCEPTfor correct denominators.View full project by clicking –>
