Data Engineer Interview Questions India 2026: Complete Guide (SQL, Spark, Kafka, dbt)

Data engineering interview question categories: (1) SQL depth ('Write a SQL query to find the second-highest salary in each department', 'Explain window functions: RANK vs DENSERANK vs ROWNUMBER; when does each produce different results?', 'Write a query using a recursive CTE to find all direct and indirect reports of a given manager in an org hierarchy table'). (2) Data pipeline design ('Design a data pipeline that ingests 1 billion events per day from 50 microservices into a data warehouse; how do you handle late-arriving events?'). (3) Apache Spark ('Explain the difference between transformations and actions in Spark', 'What is a data skew problem and how do you fix it using salting?', 'What is the difference between narrow and wide transformations?'). (4) Apache Kafka ('Explain Kafka partitioning: how are messages assigned to partitions and how do you ensure ordering within a topic?', 'What is the difference between at-least-once and exactly-once delivery in Kafka?'). (5) dbt and the modern data stack ('Explain how dbt materialises models: what is the difference between a table, view, incremental, and ephemeral materialisation?').

Data warehouse and modelling topics: Dimensional modelling (Ralph Kimball's approach: fact tables for transactional data vs dimension tables for descriptive data; star schema vs snowflake schema trade-offs; slowly changing dimensions: SCD Type 1 (overwrite), SCD Type 2 (add a new row with effective dates), SCD Type 3 (add a column for the previous value)). Column-oriented databases (why Redshift, BigQuery, and Snowflake are column-oriented and why this makes aggregation queries faster than row-oriented databases: columnar storage enables reading only the columns needed for a query). Partitioning and clustering (partitioning a BigQuery table by date column so queries that filter by date scan only the relevant partitions; clustering on additional columns for further query pruning). Data freshness trade-offs (streaming pipelines with Kafka give seconds-to-minutes freshness but are complex; batch pipelines with Spark and Airflow give minutes-to-hours freshness with simpler operations; the right choice depends on the downstream consumer's freshness requirement).

Modern data stack in Indian product companies: Ingestion (Apache Kafka for real-time event ingestion from microservices; Fivetran or Airbyte for SaaS data connector ingestion; Debezium for change data capture from Postgres or MySQL production databases). Processing (Apache Spark on Databricks or EMR for large-scale batch processing; Flink for real-time stream processing; dbt for SQL-based transformation logic in the data warehouse). Storage (Snowflake: dominant in fintech and B2B SaaS; BigQuery: dominant in companies on GCP; Redshift: dominant in companies on AWS, used extensively by Amazon India and its ecosystem). Orchestration (Apache Airflow: the dominant orchestration tool in India; knowledge of DAGs, XComs, task dependencies, and SLAs is expected). Visualisation (Metabase for open-source self-serve analytics, Superset used by Flipkart, Looker for enterprise Series C+ companies). The data lakehouse pattern (Delta Lake, Apache Iceberg, or Apache Hudi to enable ACID transactions on data lake storage) is becoming more common at sophisticated Indian data teams.

Data engineer salary ranges: junior (0-2 years): Rs 8-22 LPA. Mid-level (2-5 years): Rs 20-55 LPA. Senior (5-8 years): Rs 45-100 LPA. Staff and Principal (8+ years): Rs 90-165 LPA. Data engineers earn comparably to backend engineers at equivalent levels; senior data engineers with Spark and Kafka expertise at scale can sometimes earn more. Top-paying employers: Flipkart (one of the most sophisticated data platforms in India; Rs 35-90 LPA mid-senior), Swiggy (real-time pipelines for restaurant demand forecasting; Rs 28-75 LPA), Razorpay (financial transaction analytics at scale; Rs 30-80 LPA), CRED (member behaviour analytics; Rs 32-80 LPA), and data-first companies like CleverTap, MoEngage, and WebEngage that build data infrastructure as their core product.