The Concept
This dashboard serves as my interactive resume—a live demonstration of end-to-end BI development skills. Rather than listing competencies, I built a complete project that proves them.
The scope was intentionally comprehensive: design the data model, perform ETL transformations, write dynamic DAX measures, and deliver a polished user experience. Every aspect of the BI workflow is represented here.
Technical Implementation
Data Modeling
Star schema architecture:
- Fact table: Core metrics with foreign keys linking to dimensions
- Dimension tables: Entities for filtering and grouping (time, categories, segments)
- Relationships: Established 1:many cardinality with proper key management
WHY star schema: Optimized for Power BI’s VertiPaq engine—fast queries, intuitive relationships, and straightforward measure logic.
ETL Process (Power Query)
Data preparation workflow:
- Extract: Imported raw data from multiple sources (CSV, Excel)
- Transform:
- Standardized column names and data types
- Handled nulls and missing values
- Created calculated columns for segmentation
- Merged queries to build unified tables
- Load: Refreshed model with clean, analysis-ready data
Key learning: ETL isn’t glamorous, but it’s foundational. Clean data = reliable insights.
DAX Measures
Built dynamic calculations including:
Total metrics with context awareness:
Total Revenue = SUM(Sales[Amount])
Revenue YTD =
TOTALYTD([Total Revenue], Calendar[Date])
Time intelligence comparisons:
Revenue vs Previous Period =
VAR CurrentPeriod = [Total Revenue]
VAR PreviousPeriod = CALCULATE([Total Revenue], DATEADD(Calendar[Date], -1, YEAR))
RETURN
DIVIDE(CurrentPeriod - PreviousPeriod, PreviousPeriod, 0)
Dynamic filtering:
Active Items =
CALCULATE(
COUNTROWS(Items),
Items[Status] = "Active"
)
UX/UI Design
Navigation structure:
- Page navigation: Logical flow between overview, detail, and insights
- Bookmarks: Saved states for quick scenario comparison
- Visual hierarchy: Strategic use of size, color, and whitespace
- Consistency: Unified color palette and typography
Design principle: Data should inform, not overwhelm. Every chart serves a purpose.
Skills Demonstrated
| Area | Specific Skills |
|---|---|
| Data Modeling | Star schema, FK/PK relationships, cardinality |
| ETL | Power Query M, data cleaning, transformation logic |
| DAX | Measures, time intelligence, context manipulation, iterator functions |
| Visualization | Chart selection, visual hierarchy, color theory |
| UX/UI | Page navigation, bookmarks, user flow design |
| Tools | Power BI Desktop, Power Query Editor, DAX Studio (testing) |
What This Proves
For recruiters: I don’t just know theory—I build functioning solutions.
For clients: I understand the full BI lifecycle, from messy data to polished dashboards.
For myself: First major independent project completed. Foundation for more complex work ahead.
Reflections
What worked:
- Breaking the project into phases (model → ETL → DAX → design) prevented overwhelm
- Documenting decisions made future troubleshooting easier
- Treating it as a portfolio piece (not just practice) elevated quality
What I’d improve:
- Add drill-through pages for deeper analysis
- Implement row-level security (RLS) to demonstrate security concepts
- Create mobile-optimized layout (currently desktop-first)
Next portfolio project ideas:
- Real-time data pipeline with Azure Synapse integration
- Python-based data transformation with Databricks
- Automated reporting with Power Automate
