Adobe MTS-2 Round 1 Questions: 8 Experiences (Part 2 of 12)
- Company
- Adobe
- Reports
- 8 interview reports
- Level
- Full-time
- Stages
- Phone screen
Adobe MTS-2 round 1 in 2025–2026 was usually two coding problems with working code; caches, DP and résumé deep dives came up most.
At a glance
- Process: round 1 was a coding round, often two problems (01, 02, 05, 07, 08) or one problem in two parts (06). Candidate 02's ran on HackerRank against custom tests, 03's was 90 minutes in person, and 07's was an elimination round before two same-day rounds. Later rounds were mostly LLD (01, 04, 05, 06), then hiring manager and director.
- Asked recently: dynamic programming in 3 of 8 (01, 05, 08), caches in 2 (LFU in 02, LRU in 08), trees in 2 (01, 06), and a résumé or project deep dive in 3 (02, 03, 04).
- Difficulty: candidate 01 called both problems hard, and 07's interviewer wanted the most optimal code passing test cases. Two of the eight got offers (04, 08).
- Practise first: Least Recently Used Cache With O(1) Calls, Merge Overlapping Intervals and Search Autocomplete With Top Suggestions.
Coverage: 8 MTS-2 reports (Java backend in 02, 03, 07 and 08) from Jun 2025 to Sep 2026, in India where a place was given. The series covers 69 experienced-hire reports from Nov 2024 to Nov 2026, first technical round only; later rounds get a short note at the end of each entry. Tables count only round-one questions.
In this series
- Part 1: overview — the process, most-asked problems and a guide to every role
- Part 2 (current): MTS-2, Jun 2025–Sep 2026
- Part 3: MTS-1 and MTS-2, Nov 2024–May 2025 — 7 more MTS reports, entries 09–15
- Part 4: Computer Scientist 1, Jul 2025–Aug 2026 — 8 CS-1 reports, entries 16–23
- Part 5: Computer Scientist 1, Dec 2024–Aug 2025 — 7 more CS-1 reports, entries 24–30
- Part 6: CS-2 and Senior Computer Scientist — 4 reports, entries 31–34
- Part 7: SDE, SE2 and senior engineers — 8 reports, entries 35–42
- Part 8: C++ engineers — 9 reports, entries 43–51
- Part 9: frontend engineers — 10 reports, entries 52–61
- Part 10: Android engineers — 3 reports, entries 62–64
- Part 11: machine learning engineers — 3 reports, entries 65–67
- Part 12: release engineering and product consultant — 2 reports, entries 68–69
What came up
Every one of these rounds was hands-on coding. Three also spent real time on the candidate's own projects.
| Topic | Reports | Last seen |
|---|---|---|
| Dynamic programming | 3 of 8 | Sep 2026 (report date) |
| Résumé and project deep dive | 3 of 8 | Aug 2026 (report date) |
| Trees | 2 of 8 | Sep 2026 (report date) |
| Caches (LFU, LRU) | 2 of 8 | Aug 2026 (report date) |
| Intervals and sorting | 2 of 8 | Aug 2026 (report date) |
| CS fundamentals (OOP, SQL, messaging) | 2 of 8 | Aug 2026 (report date) |
| Matrices | 1 of 8 | Sep 2026 (report date) |
| Tries and autocomplete | 1 of 8 | Jun 2026 (report date) |
| Heaps | 1 of 8 | Nov 2025 (report date) |
| Graphs | 1 of 8 | Jun 2025 (report date) |
| Arrays and counting | 1 of 8 | Jun 2025 (report date) |
Questions we found
Algorithms / Data Structures
| Question | Reports | Last seen |
|---|---|---|
| Merge Overlapping Intervals (LeetCode 56) | 1: 02 | Aug 2026 (report date) |
| Least Frequently Used Cache Simulator (LeetCode 460) | 1: 02 | Aug 2026 (report date) |
| Sort 1 billion numbers that don't fit in memory (external sort) | 1: 03 | Jul 2026 (report date) |
| Search Autocomplete With Top Suggestions (LeetCode 642) | 1: 04 | Jun 2026 (report date) |
| Minimum Sum After K Ceiling Halvings (LeetCode 1962 family) | 1: 05 | Nov 2025 (report date) |
| 0/1 Knapsack: Best Value Within a Weight Limit (with an unreported twist) | 1: 05 | Nov 2025 (report date) |
| Message Broadcast Time in an Org Tree (parallel, then one child at a time) | 1: 06 | Nov 2025 |
| Value Filling Over Half of an Array (LeetCode 169) | 1: 07 | Jun 2025 (report date) |
| Count Pairs of Nodes That Cannot Reach Each Other (LeetCode 2316) | 1: 07 | Jun 2025 (report date) |
| Largest Product Stretch and Its Bounds (LeetCode 152) | 1: 08 | Jun 2025 (report date) |
| Least Recently Used Cache With O(1) Calls (LeetCode 146) | 1: 08 | Jun 2025 (report date) |
Low-Level Design (LLD/OOD)
| Question | Reports | Last seen |
|---|---|---|
| Refactor a Cache for Pluggable Eviction Policies | 1: 08 | Jun 2025 (report date) |
Behavioral (BQ)
| Question | Reports | Last seen |
|---|---|---|
| Justify each resume project: design choices, trade-offs, and how it would scale (practice) | 2: 03, 04 | Jul 2026 (report date) |
Language and CS Fundamentals
| Question | Reports | Last seen |
|---|---|---|
| OOP pillars, one SQL query, caching and database basics | 1: 02 | Aug 2026 (report date) |
| Where Kafka, RabbitMQ and Redis fit in a distributed system | 1: 03 | Jul 2026 (report date) |
8 MTS-2 round 1 reports
01: MTS-2 in Noida: two hard problems in one round
Report date: Sep 5, 2026 · Role: MTS-2 · Experience: 2+ years · Location: Noida, India · Result: Passed round 1, rejected after round 2
Round 1: DSA (first technical round)
- Two questions, both described by the candidate as hard. The exact prompts weren't shared.
- The first was a dynamic programming problem on a tree. Their code passed about 60–70% of the test cases.
- The second was a matrix problem. Only 10 minutes were left for it. The interviewer accepted the approach the candidate explained, but the code didn't compile and time ran out before a fix.
- The recruiter still moved them to the next round.
After round 1: Round 2 came two days later and was a low-level design of a food-court system, with extra constraints added during the interview. It was the candidate's first LLD interview. The interviewer was helpful, but the candidate felt nervous, made weak design choices, and was rejected.
02: MTS-2 Java backend: merge intervals and LFU cache on HackerRank
Report date: Aug 26, 2026 · Role: MTS-2, backend (Java) · Experience: 3 years · Result: Rejected after the director round
Round 1: DSA + Java (first technical round)
- The round ran on HackerRank. Code had to run against 6–8 custom test cases.
- Merge overlapping intervals (LC 56).
- Implement an LFU cache (LC 460), again with working code on custom tests.
- A deep dive into one point on the resume.
- Short discussion questions: OOP and its main pillars, one SQL query, caching, and databases.
The candidate felt this round went well and moved on.
After round 1: Round 2 asked for a sliding-window rate limiter, all subsequences of a string in sorted order, and Java locking. Round 3 (hiring manager) covered projects, the HLD of the current project and a URL shortener. In the director round, the candidate reversed the sentences in a file word by word with running code, then fed it to an AI tool to optimize; the interviewer wanted a stack-based version. A mirror-a-BST question followed. They were rejected the next day.
03: MTS-2 Java/Spring Boot: a 90-minute resume grilling
Report date: Jul 28, 2026 · Role: Member of Technical Staff 2 (Java, Spring Boot) · Result: Rejected after round 1
Round 1 (first technical round, in person)
- The round took about 90 minutes. All rounds had been scheduled for the same day. The candidate came from a C++ background, and the interviewer seemed to be checking whether that would carry over to Java and Spring Boot.
- Most of the time went to the resume. The candidate had to justify almost every bullet: how each project would scale, design choices and trade-offs, an old internship project, and the architecture of personal projects. One project's high-level design was drawn out (practice).
- Where Kafka, RabbitMQ and Redis each fit in a distributed system.
- Sort 1 billion numbers that don't fit in memory. The candidate worked toward external sorting and reached the expected answer.
They were rejected after this round.
04: MTS-2 offer in Noida: autocomplete with top suggestions
Report date: Jun 6, 2026 · Role: MTS-2 · Experience: close to 3 years · Location: Noida, India · Result: Offer
Round 1: DSA + project deep dive (first technical round, virtual)
- A recruiter reached out first. The round was held online.
- Build search autocomplete like a video site's search box (LC 642): match on a prefix, keep a count for each query, and return the top suggestions. The candidate used a trie that stores frequencies and top suggestions. Follow-ups: millions of queries, better ranking, lower latency, and scaling.
- A project deep dive: why the current architecture, what the trade-offs were, how it behaves at scale, and what they would improve (practice).
They passed.
After round 1: Round 2 was a rate limiter LLD and a file upload system taken from LLD to HLD. Rounds 3 (hiring manager) and 4 (director) were face to face, with AI internals, a no-library DSA problem, project architecture and failure cases. The candidate got the offer.
05: MTS-2 by referral: K halvings with a heap, then a knapsack twist
Report date: Nov 15, 2025 · Role: MTS-2 · Result: Rejected
Round 1: DSA (first technical round)
- The candidate came in through a friend's referral and a recruiter call.
- Minimize an array's sum after exactly K operations, where each operation takes one element, divides it by 2 rounding up, and puts it back. They used a max-heap: O(N log N) time, O(N) space.
- A knapsack problem with a small twist. The twist wasn't described. They solved it with 2D DP and then optimized it.
After round 1: Rounds 2 and 3 were both LLD: a configuration management service, then search autocomplete ranked by frequency with a trie, and Linux directory commands (mkdir, ls, delete) in Java, where only one function got finished. They were rejected.
06: MTS-2: one org tree, two ways to pass a message
Interview date: Nov 2025 · Role: MTS-2 · Result: Rejected after the hiring manager round
Round 1: DSA (first technical round)
- An org tree with the CEO at the root. A message goes from each parent to all its children at once. Find the minimum time for the message to reach every employee.
- Same tree, new rule: a parent tells its children one at a time (first child at t = 1, second at t = 2, and so on). Find the minimum time again.
- The candidate coded both parts, and the interviewer looked satisfied.
After round 1: Round 2 was an in-memory file system (ls, mkdir, addFile, getFile) with a follow-up on changing the file content type using Java generics; they coded a working version. Round 3, a warehouse system LLD (renting space to firms, allocation, orders), didn't go well. After a 40-minute hiring manager round, they were rejected.
07: MTS-2 Java, cut after a two-problem elimination round
Report date: Jun 9, 2025 · Role: MTS-2 (Java) · Experience: 3 years · Result: Rejected after round 1
Round 1: DSA (first technical round)
- This was a virtual elimination round. Two more rounds were planned for the same day, but only for people who cleared it.
- The interviewer wanted the most optimal time and space, and code that actually ran and passed test cases.
- First problem: Majority Element (LC 169), find the value that appears more than half the time.
- Second problem: Count Unreachable Pairs of Nodes in an undirected graph (LC 2316), count the node pairs with no path between them.
After round 1: The candidate was rejected here, so the later rounds never happened.
08: MTS-2 Java with a Node.js past: product subarray and a swappable cache
Report date: Jun 5, 2025 · Role: MTS-2 (Java; previously a Node.js developer) · Location: Noida, India · Result: Offer
Round 1: Mostly DSA (first technical round)
- Maximum Product Subarray (LC 152). The follow-up asked for the subarray itself, not just the product.
- An LRU cache in Java with O(1) get and put (LC 146).
- Then a design twist on the same code: let the user pick the eviction rule (LRU, LFU and so on) while reusing as much code as possible. That part lines up with refactoring a cache for pluggable eviction policies.
- The candidate passed.
After round 1: Round 2 covered Java OOP (abstract classes vs interfaces, polymorphism, Thread/Runnable/Callable) and a layered design from API contracts, then SQL joins. A manager round asked about projects, business impact and AI use. A director round dug into a MySQL-to-Cassandra migration. The selection call came in the third week.
Common questions
How many problems are in the Adobe MTS-2 first round?
Usually two. Five of the eight rounds had two coding problems (01, 02, 05, 07, 08) and 06 one problem in two parts; 03 had one, after a long résumé discussion, and 04 had one plus a project deep dive.
Does the Adobe MTS-2 round 1 need running code?
Often, yes. Candidate 02 ran code against 6–8 custom tests on HackerRank, and 07's interviewer wanted optimal code that passed test cases. Candidate 01's matrix code didn't compile, yet the recruiter still moved them on.
What comes after round 1 for Adobe MTS-2?
Mostly low-level design: a food court (01), a rate limiter and file upload (04), a configuration service (05), an in-memory file system and a warehouse (06). Then a hiring manager and often a director.
Keep practicing
- Practice from these reports: Merge overlapping intervals (02); LFU cache (02); mirror-a-BST (02); search autocomplete like a video site's search box (04, 05); Minimize an array's sum after exactly K operations (05); knapsack problem (05); minimum time for the message to reach every employee (06); Majority Element (07); Count Unreachable Pairs of Nodes (07); Maximum Product Subarray (08); LRU cache (08); refactoring a cache for pluggable eviction policies (08).
- Behavioral prompts from these reports: practice (03, 04).
- Adobe coding bank: every Adobe algorithm problem in the question bank (sign-in needed).
- Adobe low-level design bank: the low-level design questions.
- Adobe behavioral bank: the behavioral questions.
- Adobe overview: the Adobe interview page and Adobe hiring insights.
- More reports: the Experience feed.
In this series
Sources: Anywhere useful interview info can be found.
Comments
0No comments yet — be the first to share what you know.