I just completed Goldman Sachs’ OA. My SDE HackerRank test included the following two problems.
Description: Given a transaction value array and an integer k, count the number of consecutive segments of exactly k transactions that are strictly increasing.
First, I checked the length of the array. If it was shorter than k, I just returned 0 directly, because there’s no way to get k consecutive transactions. Then I iterated through the array: starting from the second element, I compared it with the previous one. If it was larger, I marked it as 1; if not, 0, creating a "status array".
Next, I slid a window of size k-1 over this status array. Whenever all elements in the window were 1, it meant the corresponding k transactions were strictly increasing, so I incremented the counter. The final value of the counter was the answer.
The AI I used during the interview was extremely helpful. Its best feature is that the answers it generates are completely invisible to the interviewer and undetectable by HackerRank, which is a huge advantage.
Description: Given an n×m bookshelf with encyclopedias from k authors, find the minimum number of books to select to delete all books, where selecting one deletes all same-author books in its row and column.
I handled each author separately first, since books from different authors don’t interfere with each other, and I just needed to sum up the results in the end. For a single author, I first noted down the row and column positions of all their books, then built a bipartite graph, rows and columns as vertices, and each book’s position as an edge connecting the corresponding row and column.
After that, I used the Hungarian algorithm to calculate the maximum matching of this graph. That number was the minimum number of books I needed to select for that author. Adding up the results for all authors gave the final answer.
Goldman Sachs’ OA is conducted on the HackerRank platform and must be completed within 7 days of receiving it. It includes two types of challenges. Based on the responses you provided in your application, you will be required to complete either option a or b below:
a. Programming challenge
b. Math and progrmming challenge
Programming challenge
Time allotment: 120 min
Question types: Data structures focused questions that can be completed with your choice programming language.
Math and programing challenge
Time allotment: 180 min (120 min programming + 60 min math, combined)
Question types: Combination of the above programming format and a series of math multiple choice questions.
When I started preparing for the goldman sachs hackerrank, I wanted to make sure I had the right tools. I didn’t want to waste time on random websites or outdated materials. So I focused on resources that other candidates and engineers recommended. Here’s a table that shows the ones that helped me the most:
Resource | Description |
---|---|
Linkjob | Provides mock interview practice based on a real interview question bank, along with feedback and personalized improvement plans. |
Goldman Sachs Recruitment Process Guide | Provides insights into the HackerRank test format and preparation strategies. |
HackerRank | Recommended platform for practicing coding problems, including mock assessments. |
LeetCode | Useful for solving Goldman Sachs tagged problems to enhance problem-solving skills. |
Tip: Don’t rely on just one platform. Mix up your practice to cover more ground and avoid surprises on test day.
I set up a simple weekly schedule that kept me on track without burning me out. Here’s what my routine looked like:
Monday to Wednesday: Focused on coding problems (1-2 hours each day).
Thursday: Worked on math and logic puzzles (1 hour).
Friday: Took a mock test or timed quiz to simulate real test conditions.
Weekend: Reviewed mistakes and tricky problems, then relaxed.
This schedule helped me build momentum. I didn’t try to do everything at once. Instead, I broke my prep into small, manageable chunks. That way, I could see progress every week and stay motivated.
Note: If you have less time, try shorter sessions but keep them regular. Even 30 minutes a day can make a big difference.
Time management made a huge difference for me. I didn’t want to get stuck on one question and run out of time for the rest. Before the test, I looked at how much time I should spend on each section.
I kept an eye on the clock and moved on if I got stuck. I always made sure to leave a few minutes at the end to review my answers.
To get better at managing my time, I tried a few different strategies during practice:
I took a self-assessment to see where I wasted the most time.
I tested out different strategies to see which one fit my style.
Note: Practicing with a timer helped me stay calm and focused during the real test.
Some questions on the goldman sachs hackerrank felt impossible at first. When I hit a tough one, I didn’t panic. I used a simple plan:
I read the question twice to make sure I understood it.
I broke the problem into smaller parts and solved what I could.
If I still felt stuck, I marked the question and moved on. I came back later with a fresh mind.
Sometimes, I wrote out a quick plan or pseudocode before coding. This helped me organize my thoughts and avoid silly errors. For math or logic puzzles, I drew diagrams or made lists to see patterns.
There’s also an even simpler approach: all of this can be handled with AI. Trained on countless real interview questions, the AI was a huge help for me during the interview.
Here are some common traps I fell into or saw others struggle with:
Spending too much time on one hard question can lead to rushing through easier ones at the end.
Ignoring the subjective questions initially may miss easy points.
Including too many projects on a resume can make it difficult to explain them all in detail during interviews.
Skipping practice on puzzles and basic math can result in feeling unprepared when such questions appear.
Focusing only on coding without reviewing computer science theory can cause some technical questions to be surprising.
Not practicing mock interviews early enough can make answers sound awkward and reduce confidence.
Note: Everyone makes mistakes. The key is to learn from them and adjust your approach.
To help you avoid these pitfalls, I put together a list of best practices that worked for me:
Practice a wide range of problems to build speed and accuracy. Focus on the types you feel most confident about.
Prioritize sections that can boost your score.
Only include projects on your resume that you can discuss in detail. Quality matters more than quantity.
Prepare for puzzles and review basic probability and math concepts. These often show up in the test.
Make a strong first impression with a well-prepared introduction during interviews.
Practice mock interviews to boost your confidence and improve your communication skills.
Clearly explain your thought process during coding interviews. Start with a simple solution, then improve it step by step.
Network on LinkedIn to learn about the company culture and what they look for in candidates.
Ask thoughtful questions at the end of your interview. This shows you listened and are genuinely interested.
No, Goldman Sachs typically does not share detailed numerical scores or breakdowns of the HackerRank assessment with candidates. You will only receive updates on your overall application status via email, usually within a few weeks of completing the test.
Retake policies vary by role and region, but Goldman Sachs generally restricts immediate retakes. If you are not selected, you will usually need to wait a specific period before reapplying for the same or similar roles, and thus retaking the assessment. There is no guarantee of a retake even after this period, as it depends on application availability.
The assessment balances both, but leans toward applied problem-solving tied to real-world scenarios Goldman Sachs handles. For example, questions may involve optimizing financial data processing, analyzing transaction patterns, or solving algorithmic challenges related to risk modeling, though they will still test core theoretical skills like time complexity and data structure efficiency.