Slowly Changing Dimensions (SCD) track historical changes in dimension data. Types: Type 1 (overwrite), Type 2 (full history), Type 3 (previous + current). For customer tracking, SCD Type 2 is common: add effective_date, end_date, is_current. On change: set end_date=now() and is_current=false on current row; insert new row with is_current=true. Implementation: Use MERGE in SQL or Delta Lake, or Spark with window functions to detect changes....
The complete answer continues with detailed implementation patterns, architectural trade-offs, and production-grade considerations. It covers performance optimization strategies, common pitfalls to avoid, and real-world examples from companies like Swiggy. The answer also includes follow-up discussion points that interviewers commonly explore.
Continue Reading the Full Answer
Unlock the complete expert answer with code examples, trade-offs, and pro tips - plus 1,863+ more.
Or upgrade to Platform Pro - $39
Engineers who used these answers got offers at
AmazonDatabricksSnowflakeGoogleMeta
According to DataEngPrep.tech, this is one of the most frequently asked SQL interview questions, reported at 1 company. DataEngPrep.tech maintains a curated database of 1,863+ real data engineering interview questions across 7 categories, verified by industry professionals.