Maciej Wochna portrait
Back to Work

Interactive BI Portfolio Dashboard

My first fully independent, end-to-end Power BI project demonstrating data modeling, ETL, DAX, and UX/UI design principles.

Power BIData ModelingDAXETLUX/UI
01. Context

This is my interactive CV—a comprehensive BI project I built from scratch to demonstrate proficiency across the entire data analytics workflow: from raw data to polished, user-friendly dashboard.

02. Problem

How do you prove BI competency beyond certifications? Build a real project that showcases technical depth and design thinking.

03. Constraints

First full-scale independent project—learning data modeling patterns in practice

Balancing technical complexity with user experience simplicity

No existing data infrastructure—built semantic model from ground up

04. Key Decisions

Star Schema Design

Designed fact and dimension tables following Kimball methodology. Established proper foreign/primary key relationships to enable efficient querying and cross-dimensional analysis.

ETL Implementation

Performed data transformation, cleaning, and preparation using Power Query. Handled missing values, standardized formats, and created calculated columns for analytical flexibility.

Dynamic DAX Measures

Built reusable measures with time intelligence, contextual calculations, and dynamic filtering. Leveraged CALCULATE, FILTER, and iterator functions for advanced analytics.

UX/UI Principles

Applied page navigation, bookmarks, and consistent visual hierarchy. Prioritized user flow and minimized cognitive load—data tells a story, not overwhelms.

05. Outcome

Delivered fully functional Power BI dashboard showcasing technical and design skills

Demonstrated mastery of: data modeling (star schema), ETL (Power Query), DAX (measures + time intelligence), UX/UI (navigation + bookmarks)

Created tangible proof of competency for recruiters and clients

06. Deep Dive

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:

  1. Fact table: Core metrics with foreign keys linking to dimensions
  2. Dimension tables: Entities for filtering and grouping (time, categories, segments)
  3. 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:

  1. Extract: Imported raw data from multiple sources (CSV, Excel)
  2. Transform:
    • Standardized column names and data types
    • Handled nulls and missing values
    • Created calculated columns for segmentation
    • Merged queries to build unified tables
  3. 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
07. Lessons Learned
"Building a portfolio project is different from client work. You control the scope, so resist feature creep. Focus on demonstrating core competencies clearly—technical depth matters, but so does presentation."
08. Live Dashboard