Glean Coding Interview Questions
25 Glean coding interview problems with full optimal solutions — 8 easy, 12 medium, 5 hard. Every problem ships with multiple approaches (brute-force first, then the optimal), complexity tables for each, company-specific tips on what an Glean interviewer values, and a FAQ section.
- #1easyvery frequently asked
1. Two Sum
Glean uses this as a warm-up to test whether candidates think in hash maps first — the same O(1) lookup pattern that powers their inverted-index search engine at the core.
- #3mediumvery frequently asked
3. Longest Substring Without Repeating Characters
Glean asks this to test sliding-window fluency — the same technique used in tokenizing and windowing text streams for indexing, where you need to identify the longest unique n-gram span without repetition.
- #4hardsometimes asked
4. Median of Two Sorted Arrays
Glean tests this to verify that candidates can reduce a tricky problem to binary search on the partition point — the same reasoning behind efficiently finding the rank-based cutoff in a dual-index search system without merging both indexes.
- #15mediumfrequently asked
15. 3Sum
Glean uses 3Sum to evaluate whether candidates can reduce a naive O(n³) problem to O(n²) by sorting and applying two pointers — the same space-time reasoning that shows up in multi-term query intersection optimization.
- #20easyfrequently asked
20. Valid Parentheses
Glean uses this to test stack intuition — the same mechanism that powers balanced-query parsing in enterprise search engines where unclosed brackets in a structured query must be caught before execution.
- #21easyfrequently asked
21. Merge Two Sorted Lists
Glean tests this as a gateway to Merge K Sorted Lists — a core primitive in multi-shard search result merging where ranked result sets from different index shards must be combined in order.
- #23hardvery frequently asked
23. Merge K Sorted Lists
Merge K Sorted Lists is a cornerstone Glean interview problem — it directly models merging ranked result sets from K index shards into one ordered output stream, a core operation in any distributed search engine.
- #42hardsometimes asked
42. Trapping Rain Water
Glean asks this hard-tier problem to test whether candidates can derive an O(1)-space two-pointer solution from an O(n)-space prefix/suffix scan — the same optimization mindset that matters in large-scale index traversal where memory allocation is expensive.
- #49mediumfrequently asked
49. Group Anagrams
Glean asks this because normalizing and bucketing strings by a canonical key is the foundation of term-normalization in search indexing — the same logic that maps 'ran', 'nar', and 'arn' to a single canonical form for retrieval.
- #53easyfrequently asked
53. Maximum Subarray
Glean asks this to test Kadane's algorithm — a greedy scan that mirrors how a search ranker maximizes relevance score over a contiguous window of query tokens.
- #56mediumfrequently asked
56. Merge Intervals
Glean uses this to assess interval-sweep reasoning — the same logic behind merging overlapping document time-ranges in activity timelines, or collapsing overlapping token spans in entity recognition post-processing.
- #70easysometimes asked
70. Climbing Stairs
Glean uses this to probe dynamic programming intuition — recognizing that the answer is just Fibonacci reveals whether a candidate spots optimal substructure without prompting, a skill that translates directly to ranking-function design.
- #121easyfrequently asked
121. Best Time to Buy and Sell Stock
Glean screens for greedy and sliding-window reasoning here — the same mindset used when scanning a time-series of document relevance scores to find the best retrieval window.
- #127hardsometimes asked
127. Word Ladder
Glean uses Word Ladder to test BFS on an implicit graph — a critical skill when traversing semantic neighborhoods in a word-embedding space or computing edit-distance hops between query terms for query expansion.
- #139mediumvery frequently asked
139. Word Break
Glean asks Word Break because query segmentation — splitting a raw search string like 'enterpriseaichat' into 'enterprise ai chat' — is a real pipeline step in their search engine, and the DP solution maps directly to it.
- #146mediumvery frequently asked
146. LRU Cache
Glean uses LRU Cache to test data structure composition — the same hash map + doubly-linked list pattern that sits at the heart of their real-time document caching layer, where recently accessed enterprise content must be served with sub-millisecond latency.
- #200mediumfrequently asked
200. Number of Islands
Glean uses this to probe graph traversal fluency — BFS and DFS over a 2D grid mirror the connected-component analysis used in clustering semantically related documents into topic islands.
- #206easyfrequently asked
206. Reverse Linked List
Glean tests pointer manipulation fundamentals here — the same in-place rewiring skill that matters when you're rearranging result-list nodes in a search ranking pipeline without allocating new memory.
- #207mediumfrequently asked
207. Course Schedule
Glean tests cycle detection in directed graphs here — the same topological ordering problem that arises in dependency resolution when indexing hierarchically structured enterprise knowledge bases.
- #208mediumvery frequently asked
208. Implement Trie (Prefix Tree)
Glean is an enterprise search company — Tries are the backbone of autocomplete and prefix-lookup in their search bar. Expect this to come up and expect deep follow-up questions about real-world trie extensions.
- #217easysometimes asked
217. Contains Duplicate
Glean uses this as a fast filter to test whether candidates reach for a set before a nested loop — the exact reasoning behind deduplication in search indexing pipelines.
- #238mediumfrequently asked
238. Product of Array Except Self
Glean tests prefix/suffix product reasoning here — the same divide-and-accumulate pattern used in precomputing cumulative document scores across a corpus segment without redundant recalculation.
- #322mediumfrequently asked
322. Coin Change
Glean uses Coin Change to assess unbounded-knapsack DP thinking — a pattern directly analogous to finding the minimum number of query re-expansions needed to cover a target relevance budget.
- #347mediumvery frequently asked
347. Top K Frequent Elements
Glean loves this problem because top-K selection is literally how search result ranking works — retrieving the K most relevant documents from a frequency-weighted index without sorting all candidates.
- #642hardvery frequently asked
642. Design Search Autocomplete System
This is arguably the most Glean-relevant problem that exists — it asks you to build the autocomplete system that is Glean's core product differentiator, combining a Trie with a frequency-ranked top-K retrieval engine.