Goldman Sachs Senior Software Engineer Interview Questions

1.Tell me about a time you had to work with a team outside engineering to deliver something critical. What made it challenging and how did you ensure success?

2.Tell me about someone you've mentored or coached in the last year. What was their challenge and how did you help them grow?

3.Describe a significant technical failure you experienced. What did you learn from it and how did you apply that learning to prevent similar issues later?

4.Describe a project where you had to convince multiple engineering teams to align on a technical approach. What was your strategy and how did you handle resistance?

5.Walk me through a time when you had to deliver disappointing news about a project to stakeholders. How did you handle it and what was the outcome?

6.Tell me about a time when you pushed for a significant change to an existing system or process that others were resistant to. What drove your conviction and how did you make it happen?

7.You need to migrate a legacy trading system from Oracle to PostgreSQL without any downtime. The system handles 50,000 trades per day and other applications depend on this data. What's your migration strategy?

8.You're reviewing a pull request that adds a new feature to our risk engine. The code works but uses a nested loop that becomes O(n²) when processing large portfolios. How do you handle this with the engineer?

9.Our Marcus banking platform processes loan applications that require credit decisions within 30 seconds. You notice the decision engine is making inconsistent calls for similar applicants. How would you debug this?

10.You're building a new risk calculation service that needs to process position data from our global trading desks every minute. The calculation takes 45 seconds on current hardware but needs to complete in under 30 seconds. What's your optimization approach?

11.You're working on the Marquee platform and need to aggregate real-time P&L data from 20 different trading systems, each with different data formats and update frequencies. How would you design the data pipeline to ensure sub-second latency while handling system outages?

12.Walk me through how you'd implement a rate limiter for our Transaction Banking API that serves corporate clients. Each client has different SLA tiers and the system needs to handle 100K requests per second.

13.Design a document search and discovery system for Goldman's legal and compliance teams that needs to search through 50 million documents in under 2 seconds. The system must handle legal hold requirements and maintain strict access controls.

14.We're building a new trade reporting system that needs to submit transaction reports to 12 different global regulators within their specific deadlines. Each regulator has different data formats and some require acknowledgments within 5 minutes. How would you design this system?

15.We need to rebuild the portfolio analytics engine that powers Marquee's client-facing dashboards. It currently takes 10 minutes to generate a risk report for a large hedge fund's portfolio, but clients expect sub-30-second response times. How would you architect this system?

16.Design a configuration management system for our GS Engineering Platform that serves 10,000+ internal applications. The system needs to push config changes to production within 30 seconds while ensuring no bad configurations can break critical trading systems.

17.Design the real-time settlement engine for our Transaction Banking platform that needs to process $500 billion in daily payment volume across 150 countries. How do you handle currency conversions, regulatory compliance, and ensure atomic transactions?

18.Our Marcus platform needs a fraud detection system that can analyze loan applications in real-time and make accept/reject decisions within 200ms. How would you design this system to handle 100,000 applications per day while minimizing false positives?

19.Tell me about a time you had to make a difficult tradeoff between what a client wanted and what you believed was technically sound. How did you navigate that situation?

20.Describe a situation where you had to learn a completely new domain or technology to solve a critical business problem. What was your approach and how did you ensure you didn't become a bottleneck?

21.Tell me about a time when you had to choose between shipping a feature that would help your team's metrics versus doing work that would benefit other engineering teams. How did you make that decision?

22.Walk me through a time when you had to establish technical standards or best practices across multiple teams. What was your strategy for getting adoption without formal authority?

23.Describe a situation where you discovered a significant technical debt or architectural issue that other senior engineers had dismissed or deprioritized. How did you build the case for addressing it?

24.Tell me about a time when you had to lead a cross-functional initiative involving engineering, compliance, and business stakeholders. How did you align everyone when they had conflicting priorities?

25.Estimate how many trades Goldman Sachs processes globally on a typical trading day. Walk me through your assumptions and how you'd validate this estimate.

26.A client using our Marquee platform reports that their portfolio risk calculations are showing inconsistent results when they refresh the page. How would you systematically debug this issue?

27.Our Transaction Banking platform needs to calculate foreign exchange rates for settlements. If we're off by even 0.01% on a $100 million transaction due to stale data, what's the potential cost exposure? How would you minimize this risk?

28.We're seeing a 15% increase in API timeouts on our Marcus loan origination system during peak hours. The business team says application volume hasn't changed. Walk me through how you'd investigate and resolve this.

29.Goldman processes roughly $10 trillion in trades annually. If we wanted to implement real-time fraud detection across all trading activity, estimate the infrastructure cost and architectural challenges. What's your approach?

30.You're tasked with estimating the business impact if our proprietary trading algorithms were delayed by an average of 50 milliseconds. Walk me through your analysis and what factors you'd consider.