Adobe MTS Off-Campus 2022: 7 Months at Infosys, 4 Rounds in 1 Day, Offer

TL;DR. Gagan Gupta, B.Tech CSE from Chandigarh Group of Colleges (Landran, 2021 batch), 7 months into an Infosys Specialist Programmer role, cleared Adobe's off-campus MTS loop in a single day: May 5th, 2022. Four rounds, no breaks: an online test with 32 questions (coding, aptitude, English, logical reasoning), a technical interview on BST validation and array reduction, a C++ internals grilling on red-black trees versus AVL trees and clone-linked-list-with-random-pointer, and a manager round that asked him to design a download manager. Offer letter released. Source post linked at the bottom.

This is not a "How to get into Adobe" placement guide. Those list the same 5 rounds and tell you to practice LeetCode medium. This is one specific candidate's specific off-campus loop, with the actual questions asked in each round, the actual C++ topics that consumed an entire interview, and the actual outcome for someone who had 7 months of industry experience and a competitive programming background.

The 1-day Adobe marathon

Adobe's off-campus MTS loop in 2022 compressed into a single day. The candidate received a shortlist notification on May 1st and was informed that all interview rounds would happen on May 5th. Four days of preparation time. No option to space out the rounds.

This is unusual for product companies. Most schedule rounds across 1 to 2 weeks. Adobe's single-day format is an endurance test by design. You are evaluated across DSA, language internals, system thinking, and behavioral fit, all within a 4-hour window, by different interviewers who do not share notes until the debrief. Mental fatigue is a real factor.

Who interviewed

The competitive programming background matters here. Adobe's online test includes aptitude, English, and logical reasoning sections alongside coding. Candidates from service companies often clear the coding but lose time on the aptitude sections. A CodeChef 4-star rating signals speed under timed constraints, which translates directly to the 90-minute online test format.

The 7-month tenure at Infosys is also notable. Adobe does not have a minimum-tenure requirement for off-campus MTS hires. If your technical profile passes the screening, the length of your current stint does not disqualify you. This is different from companies like Flipkart or Uber, where recruiters informally screen for at least 1.5 to 2 years before scheduling interviews.

Round 1: Online Test (90 minutes)

32 questions. 90 minutes. No negative marking mentioned. The coding section is the highest-weight component, but the non-coding sections collectively account for enough marks to swing the shortlist. Candidates who skip aptitude and English to focus entirely on coding risk falling below the cutoff.

The coding questions in this online test were described as medium difficulty, but their exact statements were not recalled in the source post. What the candidate did recall: the test platform was HackerRank, the interface was standard, and the shortlist notification came quickly after submission.

This is the first filter. Adobe's off-campus funnel receives thousands of applications. The online test is designed to reduce the pool to a manageable interview batch. If you are targeting Adobe MTS off-campus, the non-coding sections are not filler. Prepare for them.

Round 2: Technical Interview 1 (9:00 AM, ~50 minutes)

This round started with a self-introduction and a brief project walkthrough, then moved into two coding problems.

Problem 1: Minimum operations to reduce array below threshold

Statement: Given an array of numbers and a threshold k, repeatedly halve any element (choose the largest each time) until all elements are less than or equal to k. Return the minimum number of operations.

Technique: Max-heap (priority queue). Pop the max, halve it, push back. Count operations until max is within k.

Analysis: This is a direct greedy + heap problem. The trap is implementing the halving incorrectly (integer division vs. floating-point) or not recognizing that you must always target the current maximum. The time complexity is O(n log n) for the heap operations. If you attempt to sort repeatedly instead of using a heap, the solution times out.

Problem 2: Validate Binary Search Tree

Statement: Given a binary tree, determine if it is a valid BST.

Technique: Inorder traversal (should produce sorted sequence) or recursive range checking (each node must be within a valid min/max range inherited from its parent).

Analysis: This is LeetCode 98. The common mistake: checking only that left child < parent < right child, without propagating the constraint down the tree. A node deep in the left subtree must be less than all its ancestors, not just its immediate parent.

The round also included a project explanation and the question "Why are you leaving Infosys after only 7 months?" The candidate's advice: be honest. "I want to work on more technically challenging problems" is a valid and sufficient answer. Do not badmouth Infosys.

Round 3: Technical Interview 2 (10:00 AM)

This round was entirely focused on C++ language internals and advanced data structure theory. The candidate had rated themselves 8 out of 10 in C++, and the interviewer tested that rating thoroughly.

Topic 1: Red-black trees vs. AVL trees

The interviewer asked for a comparison of the two self-balancing BST variants: when you would choose one over the other, the balancing mechanics of each, and the insertion complexity guarantees.

What the interviewer was looking for: Red-black trees allow a looser balance (no subtree can be more than twice as deep as another), which means fewer rotations on insertion and deletion but slightly worse search times. AVL trees maintain a stricter balance (height difference between subtrees is at most 1), which gives better search times but more rotations on modification. In practice: red-black trees are preferred when writes are frequent (like in std::map implementations), AVL trees when reads dominate.

This is not a LeetCode question. You will not encounter this on any DSA practice platform. This is a systems/language internals question, and Adobe asks it because MTS engineers work on performance-sensitive codebases (PDF rendering, Creative Cloud infrastructure) where the choice of internal data structure has measurable latency impact.

Topic 2: AVL tree balancing and insertion complexity

Follow-up on AVL trees specifically: how rotations (LL, RR, LR, RL) work, and the proof that insertion is O(log n).

Topic 3: Compile-time vs. run-time polymorphism

Function overloading and templates (compile-time) vs. virtual functions and vtable dispatch (run-time). The interviewer wanted to know when each is appropriate and the performance implications of virtual function calls.

Topic 4: Clone a linked list with random pointers

Statement: Given a linked list where each node has a next pointer and a random pointer (which can point to any node in the list or null), create a deep copy.

Technique 1 (with extra space): Use a hash map mapping original nodes to their clones. Two passes: first create all clone nodes, then set next and random pointers using the map. O(n) time, O(n) space.

Technique 2 (without extra space): Interleave clone nodes into the original list (insert clone of node A after A, clone of B after B, etc.), then set random pointers using the interleaved structure, then separate the two lists. O(n) time, O(1) extra space.

The interviewer explicitly asked for both approaches. Knowing only the hash map solution is not sufficient for Adobe's C++ rounds.

One additional coding question

The candidate mentioned one more coding problem was asked but could not recall the specifics. This is common in multi-round single-day loops: by the third round, specific problem statements blur together in retrospective accounts.

Round 4: Manager Round (12:30 PM, ~40 minutes)

The final round was with an engineering manager. It combined behavioral assessment with one coding problem and one design question.

Behavioral segment

The manager asked about the candidate's Infosys experience, the tech stack used in their projects, and the candidate's motivation for moving to Adobe. Project details were discussed in depth, with follow-up questions on architecture decisions and the candidate's specific contributions versus team contributions.

Coding: Find unique numbers in an array

Statement: Given an array, find all numbers that appear exactly once. Constraint: most optimized approach, no modification of the original array.

Technique: Frequency map (hash map) is the standard O(n) time, O(n) space solution. The "no modification" constraint eliminates in-place approaches like negating indices. The interviewer was looking for the hash map solution articulated cleanly, with explicit mention of the space trade-off.

Design: Download Manager

Statement: Design a download manager system.

Expected scope: This is a low-level design question, not an HLD question. The interviewer wanted to see: classes (DownloadTask, DownloadManager, DownloadQueue), state machine for each download (queued, downloading, paused, completed, failed), concurrency handling (multiple simultaneous downloads), resume capability, priority management, and error handling (retry logic, timeout).

The download manager design question is distinctly Adobe. Adobe products (Creative Cloud, Acrobat Reader) ship large binary files and manage complex download states. The question tests whether the candidate can think about a problem Adobe actually cares about, not a generic "design Twitter" prompt.

The Infosys-to-Adobe pipeline

This candidate's path, 7 months at Infosys Specialist Programmer to Adobe MTS, is worth examining because it represents a specific pipeline that many engineers do not realize exists.

Infosys Specialist Programmer (SP) roles pay approximately ₹9.5 LPA in 2022. Adobe MTS roles in Bangalore pay approximately ₹22 to 28 LPA for fresh/junior hires (as of 2022-2023 cycles). That is a 2x to 3x jump in compensation for the same candidate, with 7 months separating the two data points.

The unlock was not the 7 months of Infosys experience. The unlock was the CodeChef 4-star rating and the C++ depth, both of which predated the Infosys stint. The Infosys employment served as a stability signal (the candidate is employed, has a notice period, is not desperate), but the technical assessment was entirely about DSA speed and language internals.

If you are at Infosys SP, TCS Digital, Wipro NLTH, or any service company variant that pays in the ₹7 to 12 LPA range, and your competitive programming profile or DSA depth is strong, Adobe MTS is a realistic 6-to-12-month exit. The interview tests what you already know, not what you have built at Infosys.

The actual questions, compiled

The mix tells you something: Adobe MTS interviews are not pure DSA competitions. Of the 8 assessable items, 3 are language/theory, 1 is LLD, and 4 are coding. If you prepare only LeetCode problems and skip C++ internals and LLD, you will pass 2 rounds and fail the other 2.

What this loop reveals about Adobe's hiring bar

1. C++ depth is real. If you list C++ on your resume and rate yourself 7+ out of 10, be prepared for red-black trees, virtual function internals, template metaprogramming basics, and memory model questions. This is not Java-style "what are the four OOP principles" questioning. This is systems-language-depth questioning.

2. Single-day loops reward physical stamina. Four rounds from 9 AM to 1 PM means no mental reset between rounds. Eat breakfast. Bring water. Assume the worst round will be the third one, when fatigue compounds.

3. The manager round includes a coding question. Unlike Google or Amazon where the hiring manager round is purely behavioral, Adobe's manager round tests both. Come prepared to code even in the final round.

4. 7 months of experience is enough. Adobe does not gatekeep MTS off-campus hires by tenure. If your profile passes screening, the interview evaluates your technical depth, not your resume length.

5. Off-campus is a real channel. This candidate did not go through campus placement. They applied on the careers page and got shortlisted. The competitive programming background (CodeChef 4-star) likely helped their profile stand out in the initial screening.

Source post

PapersAdda's verification standard requires a publicly accessible post URL, per-round detail from the candidate, and a stated outcome. This post meets all three criteria. The candidate named the exact questions, described each round's structure, and confirmed the MTS offer.

Source: Adobe Interview Experience, Off-Campus, 7 Months Experience by Gagan Gupta on Medium.

Verification window: The article was published on Medium and describes events from May 2022. The round-by-round detail and the specificity of the C++ questions (red-black vs AVL, clone linked list with two approaches) are consistent with Adobe's known interview patterns for MTS roles.

Read next

• Goldman Sachs Engineering Analyst 2025: The 10.5-Hour Campus Marathon, another single-day endurance loop, campus instead of off-campus
• Uber SDE-1 India 2025: Google Sheets LLD Rejection, when LLD communication, not coding ability, decides the outcome
• Swiggy SDE-2 Backend Bangalore 2025: ₹40 LPA, Rejected at Train Booking LLD, how Swiggy's hybrid rounds test DSA and design simultaneously
• Google L4 India: 7-Round Downlevel to Rejection, what happens when the recruiter says "100%" and the committee says no

How we keep this current: the patterns and figures here are compiled from public candidate reports and official notifications. Treat specifics as candidate-reported rather than official, and confirm cut-offs, dates, and eligibility on the official portal before you rely on them.