How to Ace Online Assessments at Indian Tech Companies 2026

Know what to expect before you sit the OA. Platform differences: HackerRank: runs in a browser-based editor. You can switch languages per problem. All standard library imports are available. Custom test cases can be run before submission. HackerEarth: similar to HackerRank with minor UI differences. Often used for early-stage screening. Codility: stricter time limits per problem. Has 'demo tasks' you should always run before the real assessment to understand the timing. Amazon OA (custom): includes 'Work Style Assessment' (personality/leadership principles scenarios) in addition to 2 coding problems. Complete the work style assessment calmly: there are no wrong answers if they align with Amazon's Leadership Principles. Test case visibility: most platforms show you 1-3 sample test cases publicly. You are graded against 5-20 hidden test cases. Passing all public cases does not guarantee passing all hidden cases: think about edge cases independently. Proctoring: many Indian company OAs use basic browser activity monitoring (tab switches) or webcam proctoring (CoderPad, Hackerrank's proctoring mode). Do not switch tabs during the test. Complete the OA in a quiet environment.

Poor time management is the primary reason candidates with sufficient knowledge fail OAs. The standard OA: 2 problems, 90 minutes. Allocation target: Problem 1 (typically easier): 30-40 minutes from read to full correct solution. Problem 2 (typically medium): 40-50 minutes. Buffer: 5-10 minutes for final checks. Decision tree for each problem: (1) Read the full problem. Identify: what is the input, what is the output, what are the constraints (n=10^6 rules out O(n^2), n=10^3 allows O(n^2)). (2) Identify the approach (brute force first, then optimise if time allows). (3) Write a clean solution. (4) Test against the public cases. (5) Think about edge cases: empty array, single element, all same elements, negative numbers, maximum constraints. Add edge case tests. (6) Submit. If stuck after 15 minutes on problem 2: write a brute force that handles small inputs correctly. Submit it. You get partial credit. In the remaining time, attempt to optimise. Never leave a problem blank: partial credit is always better than zero.

Knowing the most common problem types by company saves preparation time. Amazon OA: sliding window, two pointers, hash maps, binary search, graph traversal (BFS/DFS), and occasionally dynamic programming. 'Minimum time to complete K tasks', 'Maximum items packed in a bag given constraints': always think greedy and heap. Flipkart OA: similar to Amazon, with more emphasis on strings and arrays. Swiggy/Zomato OA: geography-flavoured problems (delivery routing, distance calculation), graph problems (shortest path, connected components), and simulation problems. Product company OAs in general (Medium difficulty): arrays and strings, sliding window for subarray/substring problems, two pointers for sorted array problems, hash maps for frequency/lookup problems, binary search for sorted arrays or monotonic conditions, BFS/DFS for graph/tree problems, stack for bracket matching and next greater element. What is rarely tested in OAs but is common in whiteboard: complex dynamic programming (knapsack, edit distance), segment trees, Fenwick trees. These appear in competitive programming but not typical OA rounds for SDE roles.

Being stuck in an OA is inevitable. The strategy: Do not freeze for more than 5 minutes. After 5 minutes without progress: (1) Go back to basics: what pattern does this problem resemble? Have you seen a similar problem? (2) Write the brute force solution even if it is O(n^2) or O(n^3). A working brute force passes some test cases. (3) Check the constraints again: the constraint range often hints at the expected algorithm. n ≤ 10^6 → O(n log n) or better. n ≤ 10^3 → O(n^2) is acceptable. (4) Think out loud (in your notes/comments): writing down your reasoning often unblocks you. (5) Skip to the next problem if completely blocked: come back at the end. Partial credit for problem 2 + full credit for problem 1 is a passing score at most companies. For Amazon specifically: the OA has 'additional inputs' that award extra marks for handling edge cases correctly, even if your main solution has flaws. Always test the maximum constraints.