Top 10 PwC Azure Data Factory & Azure Databricks Interview Questions (2026)
Prepare for PwC Azure Data Engineer interviews with the most frequently asked Azure Data Factory, Azure Databricks, ADLS Gen2, Delta Lake, PySpark, SQL and ETL interview questions.
Top 10 PwC Data Engineer Interview Questions
PwC Data Engineer interviews focus on SQL, Python, PySpark, Microsoft Fabric, data modeling, Medallion Architecture, dimensional modeling, and real-world data engineering scenarios. Candidates should be able to explain concepts with practical examples and business impact.
1. Tell me about yourself.
When introducing yourself in a data engineering interview, focus on your journey, technical skills, project experience, and business impact rather than only listing your resume.
Example Introduction:
"I began my journey in data by working on SQL reporting projects where I realised how powerful clean, structured information can be. Over time, I picked up Python and PySpark to handle larger datasets,
and eventually moved into BI tools like Power BI to make insights accessible to business teams. Today, I’m excited about building scalable pipelines and dimensional models that help leaders make confident decisions."2. Explain Primary Key, Surrogate Key, and Super Key.
Keys are fundamental concepts in relational databases used to maintain uniqueness and relationships between tables.
Primary Key:
• Uniquely identifies each record
• Cannot contain duplicate values
• Example: CustomerID
Surrogate Key:
• System-generated key
• Usually an auto-increment number
• Used in dimensional modeling and SCD tracking
• Example: CustomerKey
Super Key:
• Any combination of attributes that uniquely identifies a record
• Example: {Email, Phone}In real-world data warehouses, business keys identify entities while surrogate keys simplify joins and historical tracking.
3. What are the different items in Microsoft Fabric?
Microsoft Fabric is an end-to-end analytics platform that combines data engineering, data science, analytics, and reporting capabilities into one unified environment.
Microsoft Fabric Components:
Lakehouse
• Stores structured and unstructured data
Data Engineering
• Spark notebooks and data pipelines
Data Factory
• Data ingestion and orchestration
Data Science
• Machine learning development
Real-Time Analytics
• Streaming data processing
Warehouse
• SQL-based analytics
Power BI
• Reporting and dashboardsExample: An e-commerce company can ingest clickstream data into Bronze Lakehouse, clean it in Silver, create business aggregates in Gold, and visualize insights using Power BI.
4. Explain Partition By in PySpark with coding example.
Partitioning organizes data into separate folders based on column values. It improves query performance because Spark can read only required partitions instead of scanning the complete dataset.
df.write .mode("overwrite") .partitionBy("category") .parquet("/data/output")For example, storing sales data partitioned by year and month allows queries for January 2025 to scan only the required partition instead of the complete dataset.
5. Explain Medallion Architecture.
Medallion Architecture organizes data processing into three layers: Bronze, Silver, and Gold. It improves data quality, scalability, and maintainability.
Bronze Layer
-------------
Raw data from source systems
Example: CSV files from ERP
Silver Layer
-------------
Cleaned and transformed data
Example: Deduplicated customer records
Gold Layer
-------------
Business-ready aggregated data
Example: Revenue reports by region6. Write SQL query to find second highest salary without using window functions.
SELECT MAX(salary) AS SecondHighestSalary
FROM employees
WHERE salary < (
SELECT MAX(salary)
FROM employees
);The query first finds the highest salary, removes it from consideration, and then returns the maximum value from the remaining records.
7. Explain Data Structures and Algorithms in Python.
Data Structures and Algorithms help solve programming problems efficiently. In data engineering, they are useful for optimization, data processing, and problem-solving.
Common Data Structures:
• List
• Tuple
• Dictionary
• Set
• Stack
• Queue
• Linked List
• Trees
• Graphs
Common Algorithms:
• Sorting
• Searching
• Recursion
• Dynamic Programming8. Explain ROW_NUMBER(), RANK(), and DENSE_RANK().
Ranking functions are SQL window functions used to assign rankings to records based on ordering criteria.
ROW_NUMBER()
-------------
Assigns unique sequential numbers
RANK()
-------------
Same rank for ties
Leaves gaps after duplicate ranks
DENSE_RANK()
-------------
Same rank for ties
No gaps after duplicate ranksExample: If two employees have the highest salary, RANK() assigns the next employee rank 3, while DENSE_RANK() assigns the next employee rank 2.
9. Difference between Star Schema and Snowflake Schema.
Star and Snowflake schemas are dimensional modeling techniques used in data warehouses.
Star Schema:
• Central fact table
• Denormalized dimensions
• Faster queries
• BI reporting friendly
Snowflake Schema:
• Normalized dimensions
• Less data redundancy
• More complex queries
• Saves storage10. Explain Slowly Changing Dimensions (SCD) Types.
Slowly Changing Dimensions manage changes in dimension data over time while maintaining historical information.
SCD Type 1
-----------
Overwrite existing value
Example: Correcting customer name spelling
SCD Type 2
-----------
Create new record with version history
Example: Customer address changes
SCD Type 3
-----------
Store previous value in another column
Example: Current and previous city
SCD Type 4
-----------
Separate history table
Example: Current customer table + history tableInterview Tip
PwC interviewers prefer practical explanations. Always connect concepts with real project scenarios, explain why you selected a particular approach, and highlight performance, scalability, and business benefits.