ByteDance interview questions

Showing 91–120 of 160 results

• Find the Best Word for a Query SuffixHard

  This problem tests your ability to efficiently query a dictionary for the longest word ending with a specific suffix, requiring optimal data structure usage for prefix/suffix operations.

  ByteDanceSoftware EngineerCoding & AlgorithmsTechnical Screen
• Find the Earliest Contiguous Removal That Balances Index SumsMedium

  This problem tests array manipulation, prefix sums, and index tracking while handling state changes caused by subarray removal.

  ByteDanceSoftware EngineerCoding & AlgorithmsTechnical Screen
• Find the Minimum Window Containing a Multiset of CharactersMedium

  This problem tests your ability to use the two-pointer sliding window technique combined with a frequency map to identify the smallest valid substring efficiently.

  ByteDanceSoftware EngineerCoding & AlgorithmsOnsite
• Find the N-th Highest Bid in an Auction Log Without a HeapMedium

  Given an unordered list of auction bids, return the n-th highest one, treating two equal bid amounts as distinct entries, while a heap-based solution is off the table. Tests selection algorithms faster than full sorting, correct handling of duplicate values, and justifying the achieved time complexity.

  ByteDanceSoftware EngineerCoding & AlgorithmsOnsite
• Find top-paid employee per departmentMedium

  This problem tests your ability to perform group-wise filtering and ranking using window functions or joins to extract top-performing records within categories.

  ByteDanceData ScientistData Manipulation (SQL/Python)Technical Screen
• Generate combinations and permutationsMedium

  This problem tests your ability to implement recursive backtracking algorithms to generate all possible permutations and combinations of a given set of distinct integers.

  ByteDanceSite Reliability EngineerCoding & AlgorithmsTechnical Screen
• Hand-Write a Multi-Head Attention Layer: Projections, Head Splitting and MaskingMedium

  Write a multi-head attention layer from scratch: project queries, keys and values, split them into heads, apply scaled dot-product attention with optional padding or causal masks, and merge the heads. It tests tensor shape discipline, numerical stability, and the cost of attention as sequence length grows.

  ByteDanceSoftware EngineerMachine LearningOnsite
• Handle Communication Difficulty and Ambiguity on a ProjectMedium

  Use a concrete project to show how you made ambiguous decisions and communication friction visible, aligned owners and deadlines, managed residual uncertainty, and protected delivery.

  ByteDanceSoftware EngineerBehavioral & LeadershipTechnical Screen
• Handle Island Counting When the Grid Does Not Fit in MemoryHard

  This problem tests the ability to adapt graph traversal algorithms (like BFS or DFS) to scenarios where the dataset exceeds primary memory, requiring disk-based processing or tiling strategies.

  ByteDanceSoftware EngineerSoftware Engineering FundamentalsTechnical Screen
• How You Build Service Stability and Troubleshoot Production IncidentsMedium

  Tests an SRE candidate's ability to design proactively stable services and execute structured, mitigation-first production incident management under high-throughput conditions.

  ByteDanceSite Reliability EngineerSoftware Engineering FundamentalsTechnical Screen
• How do you prioritize requirementsMedium

  Explain how a TikTok platform PM generates, evaluates, prioritizes, and communicates requirements. The answer covers stakeholder intake, RICE scoring, platform leverage, technical debt, compliance, opportunity cost, internal-user metrics, and roadmap tradeoffs.

  ByteDanceProduct ManagerProduct / Decision MakingTechnical Screen
• How to deploy and tune multimodal models?Hard

  A six-part ByteDance data scientist onsite ML interview covering deploying and tuning multimodal models under GPU/VRAM, compute, and latency constraints; designing fast video retrieval over precomputed captions and embeddings with ANN and hybrid reranking; and deep-learning fundamentals — overfitting detection and mitigation, dropout and inverted-dropout scaling, normalization layers (BatchNorm/LayerNorm/GroupNorm/InstanceNorm/RMSNorm), and reinforcement learning for LLM post-training (RLHF). It tests practical systems tradeoffs alongside theoretical foundations across ML systems, retrieval, and reinforcement learning.

  ByteDanceData ScientistMachine LearningOnsite
• How to triage slow service alertsHard

  This question evaluates operational troubleshooting and incident management competencies — system observability, severity assessment, blast-radius and user-impact analysis, dependency and performance diagnostics, mitigation decision-making, and incident communication.

  ByteDanceSite Reliability EngineerSoftware Engineering FundamentalsTechnical Screen
• How would you choose a classification threshold?Medium

  This question evaluates understanding of threshold selection for probabilistic binary classifiers, testing competencies in validation-set decision threshold choice, cost-sensitive trade-offs between false positives and false negatives, effects of class imbalance, and probability calibration.

  ByteDanceData ScientistMachine LearningTechnical Screen
• Implement Expiring LRU CacheHard

  This question evaluates skill in designing and implementing efficient in-memory caching data structures, testing understanding of LRU eviction, per-entry TTL expiration, and the use of hash maps and linked lists for O(1) operations.

  ByteDanceSoftware EngineerCoding & AlgorithmsTechnical Screen
• Implement Multi-Head Attention from ScratchMedium

  An ML coding question asking you to implement multi-head attention from scratch in NumPy or PyTorch, from the query, key and value projections to the output projection. It tests tensor shape handling, scaled dot-product attention, masking, numerical stability and complexity analysis.

  ByteDanceApplied ScientistMachine LearningTechnical Screen
• Implement Sliding Windows and LRU CacheMedium

  This question evaluates algorithmic problem-solving and data-structure design, testing sliding-window and 2D submatrix reasoning for maximizing contiguous ones with bounded flips and the design of an LRU cache that enforces O(1) get/put semantics.

  ByteDanceSoftware EngineerCoding & AlgorithmsOnsite
• Implement Sorted Search and Array UpdatesMedium

  This question evaluates proficiency in array manipulation and fundamental algorithmic techniques—specifically binary search for sorted arrays, digit-wise arithmetic with carry handling, and in-place stable reordering—within the Coding & Algorithms domain and with attention to time and space complexity.

  ByteDanceSite Reliability EngineerCoding & AlgorithmsTechnical Screen
• Implement a Transformer Decoder from Scratch Without Starter CodeHard

  Implement a transformer decoder from scratch with no starter code in about 30 minutes, including embeddings, masked multi-head self-attention, feed-forward layers, residual connections and layer normalization. Tests tensor-shape fluency, causal masking and hands-on deep learning coding for a machine learning engineer interview.

  ByteDanceMachine Learning EngineerMachine LearningTechnical Screen
• Implement a binary min-heap and use it to order courses with prerequisitesHard

  A two-part coding question: implement a binary min-heap with push, pop, peek and linear-time construction without library helpers, then use it to order courses with prerequisites, always taking the smallest-numbered available course and detecting circular prerequisites. Tests heap mechanics and dependency ordering.

  ByteDanceSoftware EngineerSoftware Engineering FundamentalsTechnical Screen
• Implement an Idempotent Versioned Database UpdateMedium

  Tests your ability to design and implement a concurrency-safe, idempotent record update operation using optimistic concurrency control and deduplication keys.

  ByteDanceSoftware EngineerSoftware Engineering FundamentalsTechnical Screen
• Implement an LRU CacheMedium

  This problem tests your ability to combine a hash map and a doubly linked list to achieve constant-time performance for cache operations.

  ByteDanceSoftware EngineerCoding & AlgorithmsTechnical Screen
• Implement several OA simulation problemsEasy

  This set of problems evaluates skills in string processing, memory simulation and allocation logic, and online data-structure maintenance for dynamic contiguous segment counting, testing algorithmic design and attention to stateful invariants.

  ByteDanceData ScientistCoding & AlgorithmsOnline Assessment
• Implement stack variants and upward path sumHard

  A four-part ByteDance Data Scientist onsite coding question: implement an O(1) MinStack, a MaxStack with peekMax/popMax, a streaming median finder, and a binary-tree upward (node-to-ancestor) path-sum check. It tests data-structure augmentation, streaming order statistics, and prefix-sum tree reasoning.

  ByteDanceData ScientistCoding & AlgorithmsOnsite
• Improve TikTok ads and moderationMedium

  Answer TikTok PM strategy questions on content ecosystem comparison, less commercial ads, moderation efficiency, and AI moderation evaluation. The solution covers relevance, pacing, creative quality, risk-based queues, human oversight, precision, recall, appeals, and trust metrics.

  ByteDanceProduct ManagerProduct Design & StrategyTechnical Screen
• Infer an Alien Alphabet from Sorted WordsMedium

  This problem tests graph theory knowledge, specifically topological sorting using Kahn's algorithm or DFS, applied to dependency resolution between characters.

  ByteDanceSoftware EngineerCoding & AlgorithmsTechnical Screen
• Interpret and validate regression with interactionsHard

  This question evaluates understanding of regression modeling with interaction terms, binary outcome models and odds/marginal-effect interpretation, causal inference concepts like endogeneity and instrumental variables, model diagnostics and standard-error choices, and count-data modeling.

  ByteDanceData ScientistStatistics & MathTechnical Screen
• Intersect Two Sorted Lists of Disjoint IntervalsMedium

  Intersect two sorted lists of disjoint closed intervals in linear time, retaining one-point endpoint overlaps and advancing whichever current interval ends first.

  ByteDanceSoftware EngineerCoding & AlgorithmsTechnical Screen
• Least Frequently Used (LFU) CacheHard

  This problem tests the ability to design an O(1) time complexity cache that evicts the least frequently used items, using recency as a tie-breaker.

  ByteDanceSoftware EngineerCoding & AlgorithmsOnsite
• Maximize total catch over K hours from ponds whose yield drops after each hourMedium

  Given the current hourly catch of each fishing pond and a budget of K hours, where every hour fished at a pond lowers its catch by one, compute the maximum total number of fish. It tests modeling the declining yields, choosing an optimal schedule, and meeting a time bound better than O(K log N) when K is very large.

  ByteDanceMachine Learning EngineerCoding & AlgorithmsTechnical Screen