Power BI and SQL: Asked in Deloite

1. Challenges in Power BI Dashboard Projects and Solutions

Challenge: Slow performance when refreshing a large dataset.
Solution:
• Data Modeling: Optimize the data model by reducing the number of columns and
creating star schemas instead of snowflake schemas.
• Aggregations: Pre-aggregate data in the source or use Power BI aggregations for
faster reporting.
• DAX Optimization: Replace complex DAX measures with simpler ones, or use
calculated columns sparingly.
• Query Reduction: Disable "Auto Date/Time" and reduce the number of visuals or filters
if not needed.

Optimizing Power BI Dashboard Performance

Data Modeling
Query Reduction
Simplifying data
Minimizing
structures for
queries to
efficiency
enhance speed

DAX Optimization Aggregations

Streamlining DAX Pre-aggregating

measures for data for faster
performance reporting

Practical Use Case:

Imagine a sales dashboard handling millions of records. Instead of showing granular
transaction-level data, pre-aggregate sales data by month in the source, use summarized
data in Power BI, and refresh only recent data rather than the entire dataset.

---

2. KPI vs. Dimension

• KPI (Key Performance Indicator): A KPI is a measurable value that indicates how well
an objective is being achieved. Examples include revenue, profit margin, or customer
churn rate.
• Dimension: A dimension provides context for analyzing KPIs. It's qualitative data such
as time, region, product category, or customer segment.
Practical Example:
In a sales dashboard:
• KPI: Total Sales, Year-over-Year Growth
• Dimension: Region (North America, Europe), Time (Year, Quarter)

3. SQL Query to Find the Third Highest Salary

SELECT ESalary
FROM (SELECT ESalary, ROW_NUMBER() OVER (ORDER BY ESalary DESC) AS rank
FROM Employee) AS ranked_salaries
WHERE rank = 3;

Explanation:
We use ROW_NUMBER() to rank salaries in descending order, then select the row where the
rank is 3.

4. SQL Procedure to Select EIDs with Salary < 50,000

CREATE PROCEDURE GetEmployeesBelowSalary (@Salary INT)

AS
BEGIN
SELECT EID
FROM Employee
WHERE ESalary < @Salary;
END;

Usage:
Execute with:

EXEC GetEmployeesBelowSalary 50000;

5. Retrieve EIDs with Odd Salaries and Join with empdetails

SELECT e.EID, d.EDOB

FROM Employee e
JOIN empdetails d ON e.EID = d.EID
WHERE e.ESalary % 2 = 1;

Explanation:
We check if ESalary is odd using ESalary % 2 = 1 and join it with empdetails on EID.

6. Give Example of LAG in SQL

Using LAG for Week-over-Week Comparison

SELECT EID,
ESalary,
LAG(ESalary) OVER (PARTITION BY EID ORDER BY Week) AS PreviousWeekSalary,
ESalary - LAG(ESalary) OVER (PARTITION BY EID ORDER BY Week) AS
WeekChange
FROM EmployeeSalaries;

Explanation:
LAG retrieves the salary from the previous week for comparison. This helps analyze
week-over-week changes.

7. DAX Measure for Year-over-Year Growth

YoY Growth =
DIVIDE(
[Total Sales] - CALCULATE([Total Sales], SAMEPERIODLASTYEAR('Date'[Date])),
CALCULATE([Total Sales], SAMEPERIODLASTYEAR('Date'[Date]))
)

Explanation:
This measure compares current sales with sales from the same period last year using
SAMEPERIODLASTYEAR.

8. Unique Chart in Power BI: Decomposition Tree

Purpose:
The Decomposition Tree allows users to break down a KPI (e.g., Total Sales) into contributing
factors (e.g., by Region, Product, or Salesperson). Users can drill down dynamically to explore
hierarchies and see the impact of each category.

9. Implementing Time Intelligence in Power BI

Example: Analyze sales trends by month, quarter, or year.
Steps:
1. Create a Date Table and mark it as the Date Table in Power BI.
2. Use DAX measures for time intelligence:
 Total Sales YTD =
CALCULATE([Total Sales], DATESYTD('Date'[Date]))

3. Create slicers for Year, Quarter, and Month to enable dynamic filtering.

---

10. What is the Difference Between a Database and a Data Mart?

Database:
• Purpose: Stores raw transactional data used for day-to-day operations.
• Structure: Highly normalized to reduce redundancy (many tables, complex joins).
• Data Type: Detailed, real-time, or near real-time data such as sales transactions or
customer records.
• Use Case: Used by applications, ERP systems, or CRM systems for operational
purposes.
• Users: Operational users like clerks or application users.

Data Type
Structure
Detailed data
Highly normalized
Real-time data
Reduces redundancy
Near real-time data
Many tables
Sales transactions
Complex joins
Customer records
Database
Use Case
Users
Applications
Operational users
ERP systems
Clerks
CRM systems
Application users

Purpose

Stores raw transactional data

Day-to-day operations

Data Mart:
• Purpose: Stores summarized, subject-specific data for analytical and reporting
purposes.
• Structure: Denormalized structure (fewer tables, star schema) optimized for fast
querying.
• Data Type: Aggregated data focused on a specific business area like Sales, Marketing,
or Finance.
• Use Case: Used to generate reports and dashboards for decision-making.
• Users: Business analysts and decision-makers.
 Components of a Data Mart

Users Purpose
The primary users of the The reason for creating
data mart for decision- a data mart to store and
making analyze data

Use Case Structure

The application of the The design of the data
data mart in generating mart for efficient
reports querying

Data Type
The specific aggregated
data stored in the data
mart

————————————————————————————————————————————————
Follow me on Linkedin : https://www.linkedin.com/in/rajatjain01/

Subscribe to my Youtube Channel : https://www.youtube.com/@rajatjain001

Download Free Interview Questions PDF : https://topmate.io/rajat_jain01