Question 1

How would you handle data quality issues in a real-time ingestion pipeline?

Accepted Answer

**Strategy**: (1) Validate at source—schema registry, API validation; (2) Stream validation—Flink/Spark checks (nulls, types, ranges); (3) DLQ—quarantine invalid for triage; (4) Monitoring—alert on DLQ growth; (5) Reprocess from DLQ after fix. Don't block stream; use side-outputs; fail fast....

Question 2

Describe a fault-tolerant distributed data processing system.

Accepted Answer

**Section 1 — The Context (The 'Why')**
This system faces scale and failure challenges at production. A naive approach breaks under load, loses data, or violates compliance. The primary challenge varies by domain: notifications need preference respect; banking needs ACID; pipelines need idempotency....

Question 3

Describe the steps involved in optimizing an existing data transformation pipeline.

Accepted Answer

**Section 1 — The Context (The 'Why')**
This system faces scale and failure challenges at production. A naive approach breaks under load, loses data, or violates compliance. The primary challenge varies by domain: notifications need preference respect; banking needs ACID; pipelines need idempotency....

Question 4

Design a database schema for tracking stock trades in real-time.

Accepted Answer

**Section 1 — The Context (The 'Why')**
Real-time stock trade systems must capture every tick with millisecond latency. The primary challenge is write throughput, exactly-once semantics for P&L, and querying raw and aggregated data....

Question 5

Design an ETL pipeline to process real-time stock market data.

Accepted Answer

**Section 1 — The Context (The 'Why')**
Real-time stock market ETL must capture ticks, compute OHLCV, and feed trading systems. The primary challenge is exactly-once semantics for P&L, low latency, and handling market halts....

Question 6

Discuss data replication strategies in Kafka for fault tolerance.

Accepted Answer

**Section 1 — The Context (The 'Why')**
This system design addresses "Discuss data replication strategies in Kafka for fault tolerance...." at production scale. The primary challenges are throughput under variable load, fault tolerance across distributed components, and maintaining consistency guarantees that match business requirements. Naive monolithic approaches break when single points of failure emerge, network partitions occur, or backpressure from slow consumers cascades upstream....

Question 7

Explain the CAP theorem and its relevance in distributed systems.

Accepted Answer

**Section 1 — The Context (The 'Why')**
CAP theorem: of Consistency, Availability, Partition tolerance, only two can be guaranteed in distributed systems. The primary challenge is choosing the right trade-off for each component....

Question 8

How would you design a cost-effective data lake architecture on AWS or Azure?

Accepted Answer

**Section 1 — The Context (The 'Why')**
The primary challenge in 'How would you design a cost-effective data lake architecture on AWS or Azure?' centers on designing for production scale, correctness guarantees, and operational resilience. A naive or underspecified design fails under load: single points of failure cascade, non-idempotent operations cause duplicates on retry, and lack of observability blocks root-cause analysis....

Question 9

How would you design a data ingestion framework for heterogeneous data sources?

Accepted Answer

**Section 1 — The Context (The 'Why')**
The primary challenge in 'How would you design a data ingestion framework for heterogeneous data sources?' centers on designing for production scale, correctness guarantees, and operational resilience. A naive or underspecified design fails under load: single points of failure cascade, non-idempotent operations cause duplicates on retry, and lack of observability blocks root-cause analysis....

Question 10

How would you design a database to handle historical data storage for compliance purposes?

Accepted Answer

**Section 1 — The Context (The 'Why')**
The primary challenge in 'How would you design a database to handle historical data storage for compliance purposes?' centers on designing for production scale, correctness guarantees, and operational resilience. A naive or underspecified design fails under load: single points of failure cascade, non-idempotent operations cause duplicates on retry, and lack of observability blocks root-cause analysis....

Goldman Sachs System Design Interview Questions

Difficulty Breakdown

Key Topics Covered

How to Use This Guide

Companies asking these questions

All 10 Questions

More Interview Prep Guides

Unlock All Expert Answers