Explain your project — and actually impress the panel

1. Business problem — what question was the business trying to answer? (1–2 lines)
2. Data — sources, size, key tables/columns
3. Tools — SQL for extraction, Power Query for cleaning, Power BI for visuals, etc.
4. Your process — cleaning → modeling → analysis → dashboard (this is 60% of your answer)
5. Insights — 2–3 specific findings with numbers
6. Impact — what decision or improvement did it enable?

If you can't state the business problem in one line, the panel assumes you just followed a YouTube tutorial.

Problem: "A retail business wanted to understand sales performance, customer behaviour and product demand across regions — reporting was manual Excel files emailed every week."

Data: "Sales transactions from SQL Server (~200K rows), product and customer masters from Excel, and targets from a shared drive."

Process: "I pulled data with SQL, cleaned it in Power Query — removing duplicates, fixing data types, splitting the datetime column, and unpivoting the monthly-target sheet. I modeled it as a star schema — a central Sales fact table linked to Product, Customer and Date dimensions. Then I wrote DAX measures for Total Sales, Profit Margin and Year-over-Year growth, and designed the report with slicers and drill-through pages."

Insight & impact: "The dashboard showed 30% of revenue came from two categories, and one region was consistently missing targets due to stockouts. Management used it to rebalance inventory, and weekly reporting effort dropped from 4 hours to zero."

Structure the answer around the pipeline so it sounds complete:

• Data extraction & transformation — collected data from SQL databases, Excel and cloud storage; cleaned it in Power Query
• Data modeling — built Fact/Dimension relationships using a star schema
• DAX calculations — measures for KPIs like Total Sales, Profit Margin, YoY growth
• Dashboard design — slicers, drill-throughs, bookmarks for a clean user experience
• Deployment — published to Power BI Service, scheduled refresh, shared with stakeholders

Be ready with 5–6 specific ones and one example:

• Removing duplicates, changing data types, splitting columns (full name → first/last)
• Merging queries to combine datasets, adding custom columns (Profit = Sales − Cost)
• Unpivoting wide data into a normalized structure

Example line: "I split the DateTime column into separate Date and Time fields so filtering and time-based analysis became easier."

Name 3–4 you can actually defend: SQL Server / MySQL for structured data, Excel/CSV for offline data, SharePoint or Google Sheets for shared files, and optionally a REST API or cloud source (Azure/AWS) if you have really touched one. Never name a source you cannot answer follow-up questions about.

Pick a real, technical challenge and its solution — not "time management".

Example: "The product master had duplicate product IDs with slightly different names, which broke my relationships. I traced it in Power Query, standardized the names, and de-duplicated on ID — and I learned to always profile data before modeling it."

2–3 solid projects beat 6 shallow ones. Ideal mix: one SQL-heavy analysis project, one dashboard project (Power BI or Tableau), and one Excel/end-to-end project. Host them on GitHub, and put dashboard screenshots or a public Power BI/Tableau link on your resume.