Prepare for Tech Interviews

Hello Folks, after doing a series of Engineering Interviews for a set of graduates, I thought to write you a set of tips that will help in preparing your next Tech interview. Whether you are fresh out of university, looking to step up to a senior role, or targeting a leadership position, the good news is that tech interviews are very learnable if you know what to prepare for.

Before jumping in, here are a few interesting numbers worth knowing. According to Glassdoor, the average tech interview process takes around 23.5 days end-to-end. LinkedIn data shows that only about 1 to 2% of applicants typically receive an offer at top tier companies like Google, Meta and Amazon. A 2023 survey by Jobvite found that 73% of hiring managers ranked problem solving ability as the most important trait in a candidate - above GPA and school prestige. And according to LeetCode's annual survey, candidates who solved 100 or more problems had a pass rate of around 65%, compared to around 25% for those who solved fewer than 50.

These numbers clearly indicates that preparation is everything. So let's get into it.

1. Internships and Associates

If you are a fresh graduate or someone just starting out in the industry, interviewers are not expecting you to have built production scale systems. What they want to know is whether you can think, learn and communicate. They are evaluating your potential more than your experience.

For internship and associate roles, you can expect around 2 to 4 interview rounds. These usually include an online assessment (OA) - a timed coding challenge on platforms like HackerRank, Codility or LeetCode, usually running around 60 to 90 minutes with 2 or 3 problems to solve followed by a technical phone screen and a behavioural round. At top FAANG-tier companies, roughly 70-80% of candidates are eliminated at the OA stage alone, so this is where you need to start your preparation.

What to sharpen

Data Structures and Algorithms (DSA) is the core of every intern and associate level interview. You don't need to master everything. Focus on arrays, strings, hashmaps, stacks, queues, trees, recursion and basic graph traversal (BFS and DFS). Sorting and two pointer techniques come up very often too. Aim to solve at least 75–100 LeetCode problems before your interviews, focusing on Easy and Medium difficulty. A 2022 Blind survey found that candidates who worked through the NeetCode 150 list had a noticeably higher callback rate.

A very common mistake I see from first time interviewers is diving straight into code without saying a word. Talk through your thought process out loud. Interviewers are trained to evaluate how you think, not just whether you get to the correct answer. If you get stuck, say so and explain what you are considering. That alone shows good problem-solving instincts.

Since you have limited work experience at this level, projects and portfolios become a real differentiator. Build at least one end-to-end project you can explain from architecture to deployment. Open source contributions are a great bonus here. Some companies are also starting to ask basic system design questions even at intern level for backend roles, so it is worth having a basic understanding of REST APIs, SQL vs NoSQL databases and how a client-server architecture works.

Behavioural preparation

Use the STAR method (Situation, Task, Action, Result) for behavioural questions. Common ones at this level include things like "Tell me about a challenging project," "Describe a time you worked in a team," and "Why do you want to work here?" Prepare 3 to 5 stories from your academic or personal projects that you can adapt to different questions.

Practical tips

Start with the job description and reverse engineer your preparation around it. Practice on a plain editor or whiteboard insted of an IDE with autocomplete, since interviews won't give you that comfort. Do at least 5 to 10 mock interviews with peers or platforms like Pramp or Interviewing.io. And always time yourself. Many candidates fail not because they cannot solve the problem but because they run out of time.

2. Engineer / Senior Engineer Levels

This section is for those of you with around 2 to 8 years of experience targeting mid to senior individual contributor roles. At this stage the interview shifts from "can you code?" to "can you build real things and own them end-to-end?" The bar rises quite a bit here. Typical processes run 5 to 7 rounds and according to interviewing.io data, only about 20-25% of engineers who reach the onsite interview stage actually receive an offer. So full loop preparation matters, not just coding.

Coding expectations

You are expected to write clean and efficient code with edge cases handled. The interviewer will probe beyond your initial solution. "What is the time complexity? Can you do better? What if the input is very large?" You should be comfortable with advanced DSA topics like graphs, dynamic programming, sliding window and tries. OOP principles also come up frequently. Interviewers may ask you to design a class hierarchy or implement something like an LRU cache. For backend and infrastructure roles, basic concurrency knowledge like threads, locks and race conditions is also expected.

A 2023 Meta engineering blog noted that candidates who clearly articulated trade-offs during coding problems: for example time vs. space, or readability vs. performance, rated significantly higher on their evaluation rubric than those who simply arrived at the correct answer.

System design - the game changer

System design is honestly where senior candidates are won or lost. A typical 45 minute design round asks you to design something like a URL shortener, Twitter's feed, a distributed rate limiter, or a rid sharing backend. The framework that works is: clarify requirements → estimate scale → high-level design → deep dive → trade-offs and bottlenecks.

Key concepts you must know are load balancing, horizontal vs. vertical scaling, caching (Redis, Memcached), databases (sharding, replication, indexing), message queues (Kafka, SQS), CDNs, microservices vs. monolith, API design and CAP theorem. Resources worth investing in are System Design Interview by Alex Xu (Volumes 1 & 2) and the ByteByteGo newsletter.

Behavioural at this level

Behavioural questions go deeper here. Interviewers look for evidence of ownership, influence without authority, mentorship and handling ambiguity. Amazon's Leadership Principles are worth studying even if you are not applying to Amazon. Most major tech companies have adopted similar frameworks. Prepare stories around leading a project through ambiguity, disagreeing with a decision and the outcome, mentoring a junior team member, resolving a production incident and making a technical trade-off you stand behind.

A Karat.io study in 2023 found that senior engineering candidates who provided specific quantitative impact in their behavioural answers "reduced latency by 40%" or "saved $200K in infrastructure costs" were rated significantly higher on leadership dimensions than those who gave vague narratives. Numbers matter here.

Practical tips

Make "think out loud" a core habit. Narrate your trade-off analysis in design rounds especially. Study real engineering blogs from Netflix, Uber, Airbnb and DoorDash because these are the actual problems you'll be asked to design. Revisit OS, networking and database fundamentals since they come up as follow-up questions. And review your past projects critically. Be ready to defend every architectural decision you made.

3. Tech Lead and Above

This section is for those of you with 8 or more years of experience targeting Staff Engineer, Tech Lead, Principal Engineer or Engineering Manager roles. At this level the interview fundamentally changes in nature. You are no longer being evaluated only on technical depth. You are being assessed on your technical judgment, your organisational impact and your ability to multiply the output of the people around you.

Interview loops here are typically 6 to 8 rounds. The median time-to-hire for Staff+ roles according to a 2022 Levels.fyi report was 42 days. Nearly double that of senior engineer roles. That tells you something about how thorough the process is at this level.

Technical depth and breadth

At this level you need to demonstrate both. Interviewers will probe not just what you built, but why, and what alternatives you considered, what you'd do differently, and how it fit into the broader technical strategy. Be ready for "tell me about the most complex technical problem you've solved" and be prepared to go around 3-4 levels deep. Follow up questions like "Why did you choose Kafka over RabbitMQ?", "What were the failure modes?" and "How did you ensure observability?" will come very quickly.

System design at staff level

The problems become far more open ended. Things like "design infrastructure to support 1 billion daily active users" or "how would you migrate a monolith to microservices for a fintech company at scale?" The evaluation here is less about whether you hit every design point and more about how you navigate ambiguity, prioritise constraints and communicate a coherent technical vision. Interviewers want to see that you can drive a whiteboard session with senior stakeholders, not just answer a question. Strong candidates proactively introduce constraints, make assumptions out loud and are explicit about what they'd do in Phase 1 vs. Phase 2 vs. ideal state.

Leadership and influence

Even for technical IC tracks like Staff and Principal, you will be evaluated heavily on your ability to lead without authority. This includes driving alignment across teams, setting technical standards, sponsoring initiatives and growing other engineers. Prepare stories around defining architecture at an org level, establishing engineering practices, influencing a product or business decision through technical insight, and navigating a major failure or reorg.

According to a 2024 McKinsey report on tech talent, the most impactful tech leads demonstrated a 3:1 ratio of impact amplification. Meaning their influence on the team's output was three times their individual contribution. That is what interviewers at this level are looking for.

For Engineering Managers

If you are on the EM track, coding is lighter but not absent. You will face deep dives into how you build and maintain a high-performing team, how you handle underperformance, how you balance technical credibility with delegation and how you align engineering roadmaps with business goals. A 2023 Gartner survey found that 58% of engineers who left their company cited poor engineering management as a primary factor. So interviewers here are assessing whether you are the kind of leader who retains talent.

Practical tips

Build a personal technical portfolio narrative, a coherent story of your technical journey, the problems you have owned and the impact you have had at scale. Practice explaining complex technical topics to non-technical stakeholders in under 2 minutes. An Elegant Puzzle by Will Larson and The Staff Engineer's Path by Tanya Reilly are both worth reading for this track. Engaging in open-source work or public writing also helps demonstrate thought leadership and is increasingly referenced by interviewers at this level.

Conclusion

Regardless of which level you are preparing for, three things remain true across all of them. Preparation beats talent. Very few people fail tech interviews because they are not smart enough, most fail because they did not prepare systematically. Communication is half the score. How clearly you articulate your thinking is just as important as the thinking itself. And treat every interview as a two-way conversation. Ask smart questions and show genuine curiosity about the team's challenges.

If you are motivated enough to invest the time into the right preparation, you are already ahead of most. Keep moving forward, stay curious about upcoming tech trends and look forward to your career goals with the above points in mind. Hope these will help make your next interview a success. Cheers!

Have questions or topics you'd like me to cover in a follow-up post? Drop them in the comments below.