Microsoft

1. 1

   Coding test

   Online coding assessment, 2 questions
   HackerRank75 min2 questionsmixed

   arrays · greedy · dynamic programming · strings · tries · prefix sums · heaps · bit manipulation · hashing · palindromes

   Questions came from a pool, so sets differed (one account saw password-validity, another a substring-sum question). One easy/basic plus one harder/intermediate. No section timers, free order. 15 test cases per question, only 4 visible. Constraints reported such as n up to 2000 or 1e5.

2. 2

   Technical interview

   Technical interview round 1
   Microsoft Teams plus HackerRank (invigilation and coding)45 min

   projects · backtracking · graph traversal · DFS · linked lists · key-value store design · time and space complexity · OS · networks · threads vs processes

   Allotted 45 min, ran 45 to 60. Coding on HackerRank with no template code or input format, so assumptions and edge cases are on the candidate. Some interviewers asked CS fundamentals, others did not. About 20 min on one project.

3. 3

   Technical interview

   Second interview: technical and HR mix
   Microsoft Teams45 min

   projects · BST · AVL trees · tree BFS · celebrity problem · ML metrics · bias-variance · LLM trust and limits · cybersecurity basics · authentication · strengths and weaknesses · why Microsoft · expectations from internship

   Allotted 45 min, ran 45 to 70. Content varied a lot between candidates: some got only HR, some only DSA or technical, some a mix. One panel came from a cybersecurity domain.

Advice from candidates

49 pieces of advice, in their own words. Names removed.

• Try brute force first to pass as many test cases as possible then use the constraints to find expected time complexity and optimize.

  test · 2026-27

• Both the technical rounds were on HackerRank and questions didn’t have any template code or input format. Everything that wasn’t evident in the question was to be assumed and manually input.

  Technical interview round 1 · 2026-27

• In the interview round try not to go for the optimal solution immediately even if you know it as the interviewer wants to see how you optimize your code on the spot based on constraints. Handle edge cases very carefully as there are no premade test cases.

  Technical interview round 1 · 2026-27

• Don’t rely on the interviewer for hints as they expect you to explain the solution to them.

  interview · 2026-27

Loading the other 45…

Question patterns reported

The kind of problem, not the problem itself. Our practice questions are written to match these patterns.

• Codinggreedy: remove largest contributing gap after inserting one element
• Coding0/1 knapsack, maximise value under cost cap
• Codingpassword validity check with dictionary of banned words, trie helps
• Codingcount distinct palindromic substrings
• Codingcount substrings whose mapped digit sum divisible by length
• Codingkth largest over every growing prefix, min-heap of size k
• Codingcount index triplets with even product, parity DP, modulo
• Codinghash map or set based simple counting
• Technicalchess-piece tour via backtracking, then greedy heuristic improvement
• TechnicalDFS on adjacency list with small modification
• Technicalswap kth nodes from both ends of linked list
• Technicaltimestamped key-value store with snapshots
• Technicalbuild BST from preorder, follow-up on balancing
• Technicalfind the one person known by all but knowing none, linear
• TechnicalBFS level traversal on tree, follow-up for BST
• Onlinetrailing zeros in factorial
• Onlinefind errors in loop and array code
• Onlinedata structure with logarithmic average operations
• Codingbase conversion on binary string
• Codingprint binary tree level by level
• Codingsort list of 0s,1s,2s by pointer changes
• Codingevaluate spreadsheet cells with dependencies
• Codingfactorial without multiplication operator
• Codingcompare strings under custom alphabet order