practice a little every day, paws crossed ♡
Interview Questions
Practice 9403 coding, system design and machine learning questions from 546 companies, sorted by role, topic and interview round. Every question has step-by-step hints, a code runner that supports 8 languages and an AI mentor; members also get the full solution walkthrough.
- 9403
- Questions
- 546
- Companies
- 11
- Topics
- 8
- Languages
Showing 5551–5580 of 9403 results
- MicrosoftEasyMachine Learning EngineerExplain Transformers and deploy an LLM safely
This question evaluates understanding of Transformer architectures and practical LLM deployment competencies, covering attention mechanisms, token and positional representations, computational complexity, and production concerns like latency, cost, quality, safety, and privacy.
ML System DesignTechnical Screen - MicrosoftMediumMachine Learning EngineerExplain a project deeply
This question assesses the candidate's depth of ownership over a previous machine learning project, specifically testing their ability to articulate system design choices, technical trade-offs, and operational resilience.
Software Engineering FundamentalsOnsiteSystem DesignMLOpsArchitecture - MicrosoftMediumSoftware EngineerExplain deadlock cases and how to prevent them
This question evaluates understanding of deadlocks, concurrency control, and resource synchronization in concurrent programming, testing competency in identifying the necessary conditions for deadlock, analyzing concrete scenarios including circular waits, and considering engineering approaches for prevention, detection, and recovery.
Software Engineering FundamentalsTechnical Screen - MicrosoftMediumSoftware EngineerExplain hash map internals and edge cases
This question evaluates understanding of hash map internals and related competencies in data structures, algorithmic complexity, and concurrent programming within the Software Engineering Fundamentals domain, covering collision handling, load factor and resizing, deletion semantics, worst-case behavior, iteration order, and thread-safety concerns.
Software Engineering FundamentalsOnline Assessment - MicrosoftMediumMachine Learning EngineerExplain metrics, regularization, and ablation studies
This question evaluates understanding of classification evaluation metrics, regularization methods, and experimental design for component-level analysis in NLP and broader Machine Learning systems, testing competencies in performance measurement, regularization trade-offs, and causal attribution of model components.
Machine LearningTechnical Screen - MicrosoftHardSoftware EngineerExplain owning and debugging infra modules
This question evaluates ownership, debugging, and operational competence for low-level infrastructure components such as storage, distributed systems, and other reliability-critical infra, emphasizing reliability/performance goals, incident investigation, trade-off reasoning, and executional ownership.
Behavioral & LeadershipTechnical Screen - MicrosoftMediumSoftware EngineerExplain the Most Complex Project You Have Delivered
Explain a complex engineering project through personal ownership, constraints, alternatives, execution challenges, measurable evidence, and lessons learned.
Behavioral & LeadershipOnsite - MicrosoftMediumApplied ScientistFewest Distinct Values to Delete So at Least Half the Array Is Removed
A coding question that asks for the fewest distinct values whose complete deletion removes at least half of an integer array. It tests reasoning about how often each value occurs, proving that a choice is minimal, and the exact meaning of half when the array length is odd.
Coding & AlgorithmsOnsite - MicrosoftMediumMachine Learning EngineerFind Top K Largest Elements
This question tests the ability to efficiently find the K largest elements in an unsorted array while managing time complexity constraints through heap-based or selection-based algorithms.
Coding & AlgorithmsTechnical ScreenArraysHeapSorting - MicrosoftMediumSoftware EngineerFind Unique Zero-Sum Triplets
Find unique zero-sum triplets with distinct indices, deterministic ordering, careful duplicate suppression, and explicit time and space analysis.
Coding & AlgorithmsTechnical Screen - MicrosoftHardSoftware EngineerFind all balanced k in a permutation
This question evaluates understanding of permutations, array range reasoning, index mapping, and efficient algorithm design for detecting contiguous subarrays that exactly contain the prefix set {1..k}.
Coding & AlgorithmsTechnical Screen - MicrosoftMediumData ScientistFind common friends from directed edges
This question evaluates a candidate's ability to manipulate relational data and reason about graph reciprocity, specifically testing SQL query formulation, aggregation, deduplication, and indexing for performance in the Data Manipulation (SQL/Python) domain.
Data Manipulation (SQL/Python)Technical Screen - MicrosoftMediumSoftware EngineerFind longest substring without repeating characters
This problem tests your ability to use the sliding window technique combined with a hash map to efficiently find a maximal subarray meeting a specific constraint.
Coding & AlgorithmsTechnical ScreenSliding WindowHash MapString - MicrosoftMediumSoftware EngineerFind lowest common ancestor in tree
This question evaluates understanding of tree data structures and the lowest common ancestor concept, contrasting use of binary search tree ordering with the general binary tree case while requiring algorithmic time and space complexity reasoning in the Coding & Algorithms domain.
Coding & AlgorithmsTechnical Screen - MicrosoftMediumSoftware EngineerFind max consecutive elements with sum below target
This question evaluates algorithmic problem-solving with arrays, including reasoning about contiguous subsequences, handling sorted inputs and edge cases, and analyzing time and space complexity.
Coding & AlgorithmsOnline Assessment - MicrosoftHardSoftware EngineerFind minimum cycle cost per node
This problem tests your ability to identify cycles within a weighted directed graph and compute the ratio of cycle cost to node count efficiently.
Coding & AlgorithmsOnline AssessmentGraphBinary SearchShortest Path - MicrosoftMediumMachine Learning EngineerFind pairs with the minimum absolute difference
This question evaluates array manipulation, sorting-based reasoning, and the ability to generate ordered value pairs that reflect minimum absolute differences while handling ordering and duplicates.
Coding & AlgorithmsTechnical Screen - MicrosoftMediumSoftware EngineerFind shortest substring with n unique letters
This question evaluates proficiency in string processing, frequency management, and algorithmic optimization for locating constrained contiguous substrings. Commonly asked in the coding & algorithms domain because it reveals how candidates reason about character distinctness and efficiency trade-offs, it targets practical application of algorithmic techniques rather than purely conceptual theory.
Coding & AlgorithmsOnsite - MicrosoftMediumSoftware EngineerFind the Best Connected Strength for Every Tree Node
Compute the maximum weight connected subgraph containing each node in a tree with +1 and -1 neuron weights using rerooting dynamic programming.
Coding & AlgorithmsOnline AssessmentTree DPRerooting DPGraphs - MicrosoftMediumSoftware EngineerFind the Bugs in an AI-Generated URL Shortener
Tests code review, debugging, and refactoring skills by identifying flaws in an AI-generated in-memory URL shortener and implementing a robust, thread-safe replacement.
Software Engineering FundamentalsTechnical ScreenCode ReviewConcurrencyHashing - MicrosoftMediumMachine Learning EngineerFind the Lowest Common Ancestor in a Binary Search Tree
This problem tests the ability to leverage the Binary Search Tree (BST) property to efficiently locate the Lowest Common Ancestor (LCA) of two nodes.
Coding & AlgorithmsHR ScreenBSTTreesRecursion - MicrosoftMediumMachine Learning EngineerFind the Lowest Common Ancestor in a Binary Tree
This problem tests tree traversal techniques and recursion logic to identify the deepest node that is an ancestor to two specified nodes in an arbitrary binary tree.
Coding & AlgorithmsHR ScreenTreesRecursionDFS - MicrosoftMediumSoftware EngineerFind the Most Visited Track Position
Determine the most frequently visited track position given a series of sprint endpoints by efficiently tallying overlapping intervals.
Coding & AlgorithmsOnsitearraysprefix-sumdifference-array - MicrosoftMediumSite Reliability EngineerFind where two singly linked lists merge, given each as a sequence of node ids
Coding practice question: two singly linked lists that may merge into a shared tail are given as sequences of node ids, and you must return the id of the node where they first meet, or -1. It tests reasoning about linked list structure, linear-time traversal, and edge cases such as no shared node or one list being the other's tail.
Coding & AlgorithmsOnsite - MicrosoftMediumSoftware EngineerGreedy and Top-k Beam Search Decoding That Returns the Sentence Probability
A coding question on decoding a sequence from a next-token probability distribution: implement greedy decoding, extend it to beam search that keeps the top k partial sequences, and return the chosen sequence with its overall probability. It also probes why long sequences tend to receive low probability and how decoding should account for that.
Machine LearningTechnical Screen - MicrosoftHardSoftware EngineerGroup GPUs by node and partition by links
This problem tests your ability to model a network as a graph and perform partitioning to optimize internal connectivity constraints.
Coding & AlgorithmsTechnical ScreenGraph TheoryBipartite MatchingCombinatorial Optimization - MicrosoftMediumMachine Learning EngineerHandle Cross-Team Dependencies and Scope Conflicts
This question evaluates competency in cross-team collaboration, stakeholder management, conflict resolution, prioritization, and influencing without formal authority in a Machine Learning Engineer context.
Behavioral & LeadershipOnsite - MicrosoftEasySoftware EngineerHandling a Deadline You Will Miss, and Why Adding Engineers Late Backfires
A behavioral question about realizing that a task you own will miss its deadline and what you do next, followed by a probe on why adding more people close to delivery tends to backfire. It tests early escalation, framing options and trade-offs, ownership, and an understanding of ramp-up and coordination costs.
Behavioral & LeadershipOnsite - MicrosoftMediumData ScientistHow would you build and evaluate a classifier?
This problem tests your ability to design a binary classification pipeline from metric selection to production monitoring, with a focus on aligning technical decisions with business objectives.
Machine LearningTechnical ScreenClassificationModel EvaluationImbalanced Data - MicrosoftMediumData ScientistHow would you estimate impact without A/B?
This question tests your ability to design quasi-experimental methods to infer causal impact when randomization is not possible due to technical or business constraints.
Analytics & ExperimentationTechnical ScreenCausal InferenceQuasi-experimentsDiD