Software Engineer Interview Questions & Answers (2026)

System design

Design a URL shortener. How do you generate keys, and how do you scale reads?
Design a rate limiter. Token bucket vs sliding window — when would you pick each?
How would you design a news feed? Fan-out on write vs read?
Design a chat system. How do you handle delivery guarantees and presence?

How to approach: clarify scope and scale first (QPS, data size), sketch the high-level components, then go deep on one. Always state your assumptions out loud.

Behavioral & curveball

Tell me about a time you disagreed with a senior engineer. What happened?
Describe a production incident you caused. How did you respond?
What's the hardest bug you've ever debugged?
Why do you want to leave your current role? (Keep it forward-looking, never bitter.)
Where do you see the trade-off between shipping fast and writing clean code?

Use the STAR method for every behavioral answer — Situation, Task, Action, Result — and always quantify the result.

How interviewers actually score you

Most engineering loops grade on a rubric, not gut feel. The usual axes are problem-solving (did you reach a working, optimal solution?), coding (is it clean, correct, and tested?), communication (did you narrate trade-offs?), and collaboration (did you take hints well?). A "no hire" is rarely about a wrong answer — it's about silence, jumping to code without a plan, or ignoring edge cases. Talk through your approach before typing, and call out time and space complexity unprompted.

A 2-week prep plan

Days 1–4: drill the core patterns — two pointers, sliding window, hashing, BFS/DFS, binary search. Do 3–4 problems each and write down the pattern, not just the answer.
Days 5–8: dynamic programming and trees/graphs. These separate mid from senior candidates.
Days 9–11: system design. Pick three classic prompts and design them end to end out loud.
Days 12–14: mock interviews under time pressure. Record yourself, then watch for filler and missed edge cases.

Mistakes that quietly fail candidates

Coding before clarifying the input size, constraints, and expected output.
Not testing with an edge case (empty input, single element, duplicates).
Optimizing prematurely instead of getting a correct brute force down first.
Going quiet for two minutes — interviewers can't score thinking they can't hear.

Treat every question as a conversation. The strongest signal you can send is structured, narrated problem-solving.

Sample answers to the questions that trip people up

Reading a question is easy; saying a crisp answer under pressure is hard. Here are worked responses to three of the most common stumbling blocks.

"Reverse a linked list." Start by clarifying: singly or doubly linked, and do they want it in place? Then narrate: "I'll do it iteratively with three pointers — previous, current, and next. I walk the list once, reversing each pointer, which is O(n) time and O(1) space. The recursive version is elegant but uses O(n) stack space, so for a large list I'd prefer the iterative approach." That single answer shows correctness, complexity awareness, and a trade-off — exactly the three things the rubric rewards.

"Design a URL shortener." Don't jump to databases. Say: "Let me scope it. Roughly how many URLs per day, and what read-to-write ratio? Assuming 100M writes a day and a 100:1 read ratio, reads dominate, so I'll optimize for fast lookups." Then sketch the key generation (base62 of an auto-increment ID, or a hash with collision handling), the storage (a key-value store for the mapping), and the caching layer for hot links. Close with how you'd scale reads with a CDN or read replicas.

"Tell me about a time you disagreed with a senior engineer." Use STAR. Situation: a senior wanted to ship a feature without tests under deadline. Task: I owned quality for that service. Action: I proposed a thin test layer covering the critical path only, so we shipped on time with a safety net, and I framed it as risk reduction, not process for its own sake. Result: we caught a regression in staging that would have hit production. Keep it forward-looking and never bitter.

Round-by-round: what a typical loop looks like

Most companies run a recruiter screen, a technical phone screen, and a 4–6 hour onsite (often virtual). The phone screen is usually one medium coding problem. The onsite splits into two coding rounds, one or two system design rounds (heavier as you get more senior), and a behavioral or "values" round. Some companies add a domain round — frontend, backend, or infrastructure depth.

Knowing the structure helps you allocate prep. Junior candidates should weight coding heavily. Mid-level candidates need solid coding plus at least one strong system design story. Senior and staff candidates are judged mostly on design, scope, and the behavioral round, where leadership and influence matter more than raw algorithm speed.

What separates a strong answer from an average one

An average candidate writes correct code silently. A strong candidate states assumptions, names the brute-force solution, explains why they're optimizing, writes clean code with good names, tests it on an edge case, and states the complexity without being asked. The code is often similar; the communication is night and day.

The same applies to design. Average candidates list components. Strong candidates state the scale, justify each choice against that scale, and proactively call out bottlenecks and how they'd address them. They also know when to say "I'd reach for an existing tool here rather than build it."

How expectations shift by company and level

At large product companies, expect a heavy algorithms bar and structured rubrics. At early-stage startups, expect more practical, build-something rounds and questions about shipping under ambiguity. At infrastructure-heavy companies, expect depth on concurrency, systems, and reliability.

Level changes the bar more than company. A new grad who reaches an optimal solution with hints passes. A senior engineer is expected to drive the problem, anticipate edge cases, and reason about production concerns — monitoring, rollout, failure modes — without prompting.

Frequently asked questions

How long should I prepare? For a focused candidate, four to six weeks of consistent daily practice is typical. If you're rusty on data structures, give yourself eight.

Do I need to memorize solutions? No. Memorizing fails the moment a question is slightly different. Learn the underlying patterns — sliding window, two pointers, BFS/DFS, dynamic programming — so you can adapt.

What if I get stuck in the interview? Say what you're thinking, state the part that's blocking you, and propose a brute-force fallback. Interviewers give hints to candidates who are clearly engaged. Silence is the only real failure.

How important is the behavioral round? More than most engineers assume. A strong technical performance with a weak behavioral round can still be a no-hire at companies that screen hard for collaboration.

An expanded question bank by theme

Once you've covered the headline questions, broaden your reps with these. Each is a real, frequently asked prompt — practice saying the approach out loud, not just solving on paper.

Arrays and strings: Find the longest substring without repeating characters. Merge two sorted arrays in place. Rotate an array by k positions with O(1) extra space. Group anagrams together. Find the majority element. Implement string-to-integer with edge cases.

Trees and graphs: Validate a binary search tree. Find the lowest common ancestor of two nodes. Serialize and deserialize a binary tree. Detect a cycle in a directed graph. Find the number of islands in a grid. Clone a graph.

Dynamic programming: Climbing stairs and coin change. Longest common subsequence. Edit distance between two strings. Maximum subarray sum. House robber. Word break.

Concurrency and systems (mid to senior): What's a race condition, and how do you prevent one? Explain a deadlock and the four conditions for it. How would you design a thread-safe counter? What's the difference between a process and a thread?

Follow-up questions interviewers love

Strong interviewers rarely stop at the first answer. After you solve a problem, expect: "Can you reduce the space complexity?" "What if the input doesn't fit in memory?" "How would you test this?" "What happens with malicious or malformed input?" "How would this change if the data were streaming instead of static?" Prepare for the follow-up, because the initial solution is often just the entry ticket.

For design, the follow-ups push on scale: "Now it's 10x the traffic — what breaks first?" "How do you handle a regional outage?" "Where's your single point of failure?" Anticipating these and addressing one or two proactively is the fastest way to signal seniority.

A realistic two-week study plan

Days 1–3: Review core data structures and their operations' complexities. Do two problems each on arrays, strings, and hashing. Write down the pattern behind each.
Days 4–6: Trees, graphs, and recursion. These appear in most onsites. Do BFS and DFS until they're muscle memory.
Days 7–9: Dynamic programming and a refresher on sorting and searching. DP is the most common reason mid-level candidates stall.
Days 10–11: System design. Work through three classic prompts end to end, out loud, sketching components and naming bottlenecks.
Days 12–13: Mock interviews under time pressure, ideally with another person or a tool that gives feedback. Record yourself and review for silence and missed edge cases.
Day 14: Light review of your weakest area and your behavioral stories. Rest — fatigue costs more points than one extra problem earns.

The day before and the day of

The night before, stop grinding new problems; review patterns and your behavioral stories instead. Confirm the format, the tools, and the time. On the day, read each problem twice, restate it to confirm understanding, and think out loud from the first second. Treat the interviewer as a teammate solving the problem with you, because that collaboration signal is often the deciding factor between two technically similar candidates.

How to turn this question list into real readiness

A list of questions is raw material, not preparation. The candidates who convert practice deliberately, and the method is the same regardless of role: focus on coding fluency, design reasoning, and clear narration.

Start by answering out loud, never silently. Comprehension and recall under pressure are different skills, and only spoken practice builds the second. Record yourself so you can hear the filler words, the hedging, and the moments where your structure falls apart — things you never notice while speaking.

Then score yourself against a simple rubric: was the answer structured, specific, and relevant to what was asked? Did it land on a concrete result or trade-off? Rebuild the weakest answers and run them again. A useful daily rep is to solve a problem while talking, then re-solve it silently for speed.

Use spaced repetition rather than a single cram. Three short sessions across a week beat one long session the night before, because the goal is durable recall under stress, not short-term familiarity. Finally, simulate pressure with at least two timed mock interviews before the real thing — pressure changes how you think, and you want to have felt it before it counts.

A final pre-interview checklist

Run through this the day before:

Can you state time and space complexity for every solution without being asked?
Do you test each solution on an empty, single-element, and duplicate-heavy input?
Can you design one classic system end to end and name its bottleneck?
Do you have three behavioral stories that show collaboration and ownership?
Have you researched the company, the team, and the specific role enough to tailor your answers and ask sharp questions of your own?
Have you prepared two or three genuine questions to ask the interviewer that show you understand the role?

If you can answer yes to each, you're ready. Get a good night's sleep — being rested will do more for your performance than one more hour of practice.

Practice these questions with AI

Reading questions is step one. The candidates who convert are the ones who rehearse out loud and iterate on feedback. Paste your target job description into ClavePrep to generate role-specific questions, run a free AI mock interview (text or voice), and get structured feedback on each answer. Build your behavioral stories first with the free STAR Answer Builder.

Coding & data structures