Goldman Sachs Interview Questions 2026 — Engineering Analyst & Fresher Hiring Guide
Goldman Sachs hires engineering freshers in India through campus drives, its Engineering Campus Hiring Program and the India Hackathon. Goldman Sachs describes a HackerRank assessment, a short video interview and final-round interviews with engineers; candidates commonly report questions such as sliding-window maximum and other DSA problems, plus CS fundamentals like ACID properties, normalisation and process vs thread.
Preparing for your Goldman Sachs Interview?
Practice a simulated AI Mock Interview using real Goldman Sachs questions and receive instant feedback.
Goldman Sachs is one of the most sought-after finance employers for Indian engineering graduates. Most freshers join as analysts in engineering, building trading, risk, data and platform systems from the firm’s offices in Bengaluru and Hyderabad. Non-engineering students also apply to analyst programmes in other divisions.
This guide separates what Goldman Sachs states officially from what candidates commonly report, so you prepare for the process as it actually runs.
How Goldman Sachs Hires Freshers in India
| Route | Who it is for | What Goldman Sachs states |
|---|---|---|
| New Analyst Program (India) | Final-year bachelor’s and master’s students | Programme page lists students graduating in the stated year; locations shown in the application |
| Summer Analyst Program (India) | Penultimate-year students | Internship programme; the 2027 programme page targets students graduating in 2028 |
| Engineering Campus Hiring Program | Engineering students across streams | 2025-intake FAQ: virtual on Zoom and HackerRank, one-hour proctored test with camera on, software engineer role in Bengaluru or Hyderabad |
| Goldman Sachs India Hackathon | Engineering students graduating in 2027 and 2028 (2026 edition), listed colleges | 12-hour online hackathon with CS and Quant tracks; top performers fast-tracked to a pre-placement interview |
The selection stages
| Stage | What Goldman Sachs says | What candidates commonly report (India) |
|---|---|---|
| 1. Online assessment | Engineering applicants take a HackerRank assessment and can choose programming or programming-and-maths | 2–3 coding problems, easy to medium; sometimes maths or CS MCQs |
| 2. Video interview | Selected applicants complete a video interview of about 30 minutes | Short recorded behavioural questions; not reported in every drive |
| 3. Final rounds | Interviews with three to four engineers at junior and senior levels | 2–4 technical rounds (DSA, CS fundamentals, projects, light design), then a managerial or HR round |
Exact counts vary by campus, year and division. Your invite email is the source of truth.
🚀 Start Free Goldman Sachs Mock Interview
Practice Goldman Sachs DSA, CS fundamentals and project deep-dive questions with real-time AI feedback on clarity and structure.
Start Free Mock InterviewTechnical Interview Round: Key Focus Areas
1. Data Structures and Algorithms
- Commonly reported topics: sliding windows, arrays and strings, hashing, linked lists, stacks and queues, trees, graph traversal and topological sort.
- How to prepare: solve 80–120 medium problems in one language. For each, state brute force, optimise, then code and test edge cases.
2. CS Fundamentals
- Commonly reported topics: OOP and polymorphism, abstract classes vs interfaces, normalisation, ACID, processes vs threads, deadlock, operating system basics.
- How to prepare: write one-paragraph answers with a real example for each topic, then say them aloud.
Check whether your resume has the keywords Goldman Sachs' screening looks for. Scan Resume with ATS Checker →
3. Projects and Design
- Expect to explain your project’s architecture and database schema, and to sketch a simple system such as an auction or booking service with attention to concurrency.
4. Puzzles and Quantitative Reasoning
- Puzzles (egg drop, weighing, probability) are commonly reported. Practise reasoning aloud from small cases rather than memorising answers.
6-Week Goldman Sachs Preparation Plan
- Weeks 1–2 — DSA core: arrays, strings, hashing, two pointers, sliding window, stacks/queues. Two timed HackerRank-style mocks per week.
- Week 3 — Advanced DSA: trees, graphs (BFS/DFS, topological sort), heaps, basic dynamic programming.
- Week 4 — Quant and CS: probability, permutations and puzzles; DBMS (ACID, normalisation, indexing, SQL), OS (threads, deadlock), OOP in your main language.
- Week 5 — Projects and design: a 2-minute and a 10-minute version of each project; practise two small design problems with schemas.
- Week 6 — Interviews: three full mock interviews, a video-interview rehearsal, and a specific “why Goldman Sachs” answer.
Read our technical interview mistakes developers make and STAR method examples for IT interviews before the final rounds.
Practice communication, confidence, pacing, filler words, and HR responses. Try Goldman Sachs HR Mock Interview →
Salary and Offer Details
Goldman Sachs’ engineering programme FAQ said compensation details would be shared later in the process, and the figures circulating online are self-reported and vary widely by role and year, so we have not quoted a number here. Use your offer letter as the reference and convert CTC to monthly take-home with our CTC to in-hand salary calculator.
Resume Tips for Goldman Sachs
- Format: one page, single column, no tables or text boxes. Build your resume in an ATS-safe template and check your ATS score before applying.
- Keywords: Data Structures, Algorithms, Java or C++ or Python, SQL, OOP, Multithreading, REST APIs, Git, Linux, competitive programming ratings if strong.
- Templates: start from our Software Engineer Resume India or Java Developer Resume India examples.
- Keywords Map: cross-check with the Software Engineer ATS Keywords sheet.
- Company Hub: see the Goldman Sachs Company Hub, and compare with the JPMorgan Chase interview guide if you are applying to both. Browse live openings on the FundoCareer fresher jobs board.
Sources
Hiring processes change between drives. These are the sources used for this guide (checked November 2026):
- Goldman Sachs — New Analyst Program (India): goldmansachs.com
- Goldman Sachs — Summer Analyst Program (India): goldmansachs.com
- Goldman Sachs — Engineering Analyst Campus Hiring Program FAQ (2025 intake): goldmansachs.com
- Goldman Sachs — India Hackathon 2026: goldmansachs.com
- Goldman Sachs — Careers blog, guide to HackerRank: goldmansachs.com
- Goldman Sachs — Students, Prepare: goldmansachs.com
- GeeksforGeeks — Goldman Sachs Engineering Campus Hiring Program interview experience (analyst): geeksforgeeks.org
Process details marked “commonly reported” come from candidate accounts, not from Goldman Sachs, and can differ for your drive.
Goldman Sachs Interview Questions with Model Answers
These are real questions asked in Goldman Sachs interviews in India, with model answers that interviewers have told us they score highly. Each answer is self-contained.
The brute-force solution scans each window, costing O(n·k). The optimal approach uses a double-ended queue (deque) that stores indices in decreasing order of their values. For each index i: 1) pop from the front if that index has left the window (index <= i - k); 2) pop from the back while the value at the back is less than or equal to arr[i], because those elements can never be a window maximum again; 3) push i to the back; 4) once i >= k - 1, the front of the deque is the maximum of the current window, so append arr[front] to the result. Each index is pushed and popped at most once, so the total time is O(n) with O(k) extra space.
A candidate write-up for Goldman Sachs' engineering campus programme reports exactly this problem in round one. Explain why the deque stays monotonic before you start typing.
ACID describes the guarantees a transactional database gives. Atomicity: a transaction completes fully or not at all — a fund transfer cannot debit one account without crediting the other. Consistency: every transaction moves the database from one valid state to another, respecting constraints such as non-negative balances. Isolation: concurrent transactions do not see each other's intermediate states; isolation levels (read committed, repeatable read, serializable) trade strictness for throughput. Durability: once committed, data survives crashes, typically via write-ahead logs. In finance, a lost or half-applied transaction is a real monetary error, which is why banks rely on these guarantees.
Give the transfer example and name at least one isolation anomaly (dirty read, non-repeatable read or phantom read) to stand out.
A process has its own memory space and resources; threads inside a process share the heap, code and open files but have separate stacks and registers. Threads are cheaper to create and switch, and communicate through shared memory, but need synchronisation to avoid race conditions. Prefer multiple processes when you need isolation — a crash or memory corruption in one should not take down the others — or when a language runtime limits true parallelism of threads. Prefer threads for lightweight concurrency inside one service, such as handling many I/O-bound requests.
Candidates report follow-ups on deadlock. Be ready with the four Coffman conditions and one way to prevent deadlock, such as acquiring locks in a fixed global order.
Normalisation organises tables to reduce redundancy and update anomalies. 1NF removes repeating groups so each column holds atomic values; 2NF removes partial dependencies on part of a composite key; 3NF removes transitive dependencies, so non-key columns depend only on the key. You might denormalise for read-heavy reporting or analytics workloads where joins become a performance bottleneck — for example, storing a pre-computed daily position summary — accepting extra storage and the burden of keeping copies consistent.
Tie it to an example from your own project schema. Interviewers often move straight from theory to 'design the tables for your project'.
Start by clarifying requirements: users list items, other users place bids until an end time, and the highest valid bid wins. Core tables: users(user_id, name, email); items(item_id, seller_id → users, title, description, start_price, end_time, status); bids(bid_id, item_id → items, bidder_id → users, amount, created_at). Add an index on bids(item_id, amount DESC) to fetch the current highest bid quickly. Discuss concurrency: two bids arriving at the same moment must not both 'win' — use a transaction that checks the current highest bid and inserts atomically, or optimistic locking with a version column on items. Close the auction with a scheduled job that marks the item sold and records the winning bid.
This open-ended question is reported in a Goldman Sachs campus write-up. There is no single right answer — structure (requirements, entities, relationships, concurrency) matters more than completeness.
Use Floyd's tortoise-and-hare algorithm. Move a slow pointer one step and a fast pointer two steps at a time; if they ever meet, there is a cycle, and if fast reaches null, there is none. To find the start, reset one pointer to the head and move both one step at a time; the node where they meet again is the start of the cycle. This works because the distance from the head to the cycle start equals the distance from the meeting point to the cycle start (modulo the cycle length). Time O(n), space O(1).
Mention the hash-set alternative (O(n) space) first, then improve it. Showing the trade-off is what panels look for.
An abstract class can hold state (fields), constructors, and both abstract and concrete methods; a class can extend only one. An interface defines a contract; since Java 8 it can contain default and static methods, and since Java 9 private methods, but it cannot hold instance state, and a class can implement many interfaces. Use an abstract class when related classes share state and behaviour (a base Account with a balance field). Use an interface to describe a capability shared by unrelated classes (Comparable, Runnable).
Candidate reports show Java and OOP questions in later rounds. Add a one-line example from your own code if you can.
The answer is 14. If you drop the first egg from floor x and it breaks, you must check floors 1 to x - 1 one by one with the second egg, so x drops in the worst case. To keep the worst case constant, each later first-egg drop should go up by one fewer floor: x, x + (x - 1), x + (x - 1) + (x - 2), and so on. You need x + (x - 1) + ... + 1 >= 100, i.e. x(x + 1)/2 >= 100, which gives x = 14. So you drop from floors 14, 27, 39, 50, 60, 69, 77, 84, 90, 95, 99, 100.
Puzzles are commonly reported in Goldman Sachs technical rounds. Think aloud — interviewers score the reasoning, and a correct number with no derivation earns little.
Use a tight structure. 1) Context in one sentence: what the project does and who used it. 2) The specific challenge: for example, slow API responses under load, a race condition, or an ML model overfitting. 3) What you tried, including what did not work. 4) The fix and the measurable result — 'adding a composite index cut the query from 1.2 s to 80 ms'. 5) What you would do differently now. Be ready to draw the architecture and the database schema, because follow-up questions usually go one level deeper than your resume.
Managerial rounds in candidate reports focus on obstacles, motivation for personal projects and the final-year project. Own your contribution clearly — say 'I', not 'we', for your part.
Be specific. Explain that engineering at Goldman Sachs builds systems where correctness, latency and reliability carry direct financial consequences — trading, risk, payments and data platforms. Mention one concrete thing you engaged with, such as the India Hackathon, a Goldman Sachs engineering blog post, or a public open-source project from the firm, and what you took from it. Then connect your strengths — say, strong DSA and an interest in data-intensive systems — to the team or track you are interviewing for, and say what you want to learn in your first year.
Avoid compensation or brand-name answers. One genuine, verifiable detail is more convincing than general praise.
Common Mistakes in Goldman Sachs Interviews
These are the mistakes that eliminate candidates at this stage — often before the technical round even begins.
- Preparing only coding and skipping maths and aptitude — the HackerRank assessment can include a programming-and-maths option.
- Jumping to code without stating the brute-force approach, the optimal approach and their complexities.
- Giving textbook definitions of ACID or normalisation without a banking or project example.
- Freezing on puzzles instead of reasoning aloud from small cases.
- Listing technologies or projects on the resume that you cannot defend in a deep-dive round.
Frequently Asked Questions
Ready to Land Your Offer at Goldman Sachs?
Practice with real interview questions and optimize your resume using FundoCareer's placement prep suite.