Round 1 – Technical Round (Data Structures, SQL, Spark)

This round focused on testing core technical knowledge, logical thinking, and hands-on understanding of Spark and SQL. Candidates were allowed to choose any programming language they were comfortable with. Coding questions primarily revolved around lists, dictionaries, sets, and tuples.

Spark Concepts and Questions

➤ Why is Spark faster than Hadoop?

➤ RDD vs DataFrame vs Dataset

➤ Components of the Apache Spark ecosystem

➤ Fault tolerance and lazy evaluation in Spark

➤ Types of transformations in Spark

➤ Difference between map and flatMap

➤ Persist vs cache

➤ Repartition vs coalesce

➤ Types of shared variables in Spark

➤ Runtime architecture of Spark

SQL Concepts and Questions

Topics included joins, nested queries, common table expressions, and window functions.

Specific questions asked:

➤ What is the difference between left semi join and left anti join?

➤ How are null values handled in joins?

➤ Why are window functions used?

➤ How does partitioning help in speeding up queries?

This round emphasized clarity of fundamentals rather than complex problem-solving.

Round 2 – Managerial Round (Spark Architecture and Azure)

This round was heavily scenario-based and focused on architecture-level understanding. Candidates were expected to give precise and structured answers and were encouraged to clarify assumptions before responding.

Spark Architecture and Fault Tolerance

➤ Explain Spark architecture and its components

➤ How does each Spark component contribute to fault tolerance?

➤ What happens if a Spark job fails at 40 percent completion?

➤ How is reprocessing handled in case of job failure?

Azure Scenario-Based Questions

Questions were primarily around Azure Synapse, Azure Data Factory, and Azure Databricks.

➤ What is Azure Synapse?

➤ How are failed jobs handled and reprocessed in Azure pipelines?

➤ How would you migrate on-premise data to the cloud?

➤ What are the different activities in Azure Synapse and when should each be used?

➤ What optimization techniques would you apply in Azure-based data pipelines?

This round tested real-world experience and the ability to reason through failure and recovery scenarios.

Round 3 – Partner and HR Round

The final round focused on behavioral aspects, stress handling, team management, and long-term alignment with the organization. The partner also explained team structure, role expectations, and future opportunities.

Behavioral and HR Questions

→ What was your most challenging task in your previous role?

→ Why are you leaving your current job?

→ What are your expectations from our company?

→ How do you see your growth over the next five years?

→ What do you expect from the company to support your growth?

→ What are your expectations regarding work-life balance?

→ Candidates were encouraged to ask questions to clarify role responsibilities, team expectations, and the technologies they would be working with.

Final Thoughts

Big4 Data Engineer interviews focus strongly on fundamentals, real-world scenarios, and architectural clarity rather than trick questions. A solid understanding of Spark internals, SQL concepts, and Azure services is essential. Clear communication, structured thinking, and asking the right clarifying questions play a crucial role in succeeding across these rounds.

This interview experience highlights the importance of being well-prepared on both technical depth and professional expectations when interviewing with Deloitte and EY.