Autodesk interview process (2026)
Round-by-round breakdown of the Autodesk software engineering interview across 4 levels and 18 rounds: what each round covers, what interviewers evaluate, sample questions, and preparation tips.
SDE Intern (on-campus, Bengaluru / Pune)
Difficulty: Medium · Timeline: ~1-2 weeks (campus drive) · Recommended prep: 4-6 weeks
-
Round 0: Online Test (CodeSignal / campus platform)
Assessment · ~60-90 min · difficulty 2/5 · elimination round
Focus: Strings, arrays, sliding window, hashing, basic DP
Campus drives open with a timed online test of 3-4 coding problems, typically on CodeSignal or the campus platform. Reported problems lean on strings, arrays and sliding-window patterns at LeetCode easy-medium level. This is a hard cut: in one on-campus GeeksforGeeks report only ~15 students out of a large pool advanced, and a fast finisher cleared all problems in 30 minutes.
What they evaluate: Speed and accuracy on easy-medium DSA under time pressure.
Sample question: Longest Substring Without Repeating Characters - given a string, return the length of the longest substring in which no character repeats. Input: the string on one line. Output: the length.
Tips: Aim to finish every problem, not just pass partial cases - the cut after this round is steep. Sliding-window and hashing templates cover most reported questions.
-
Round 1: Technical Interview 1 (DSA + CS fundamentals)
Onsite · ~60 min · difficulty 2/5 · elimination round
Focus: Linked lists, trees, binary search, OOPs, DBMS, OS, resume walk-through
Starts with your intro, resume and any internship work, then a DSA problem coded in a shared document - a reported question was reversing a linked list in groups of K, with the interviewer probing the recursive approach and complexity. Expect follow-up theory from OOPs, DBMS and OS. In the on-campus report, 30 students entered this round from 80 applicants.
What they evaluate: Clean pointer/recursion handling, complexity analysis, and grasp of core CS subjects.
Sample question: Reverse a linked list in groups of K - given K and the list values, reverse every consecutive group of K nodes; a trailing group smaller than K stays as-is. Input: K, then the values. Output: the resulting list.
Tips: Talk through the recursive and iterative versions before coding; interviewers here explicitly reward thinking aloud.
-
Round 2: Technical Interview 2 (pure DSA)
Onsite · ~60 min · difficulty 3/5 · elimination round
Focus: Hashing, grids/graphs, medium LeetCode patterns, communication
A problem-solving round with two medium questions - Group Anagrams and counting the Number of Islands were reported, the latter with only ~10 minutes to code. The interviewer pushes you to narrate your approach continuously. Another steep filter: 5 of 30 candidates advanced in the on-campus drive.
What they evaluate: Pattern recognition on medium DSA and the ability to code fast while explaining.
Sample question: Number of Islands - given an R x C grid of 0s and 1s, count the connected groups of 1s (4-directional adjacency). Walk through your DFS/BFS choice and its complexity.
Tips: Practice grid DFS/BFS until you can write it in under 10 minutes; that exact time box was reported.
-
Round 3: HR Round
Behavioral · ~30 min · difficulty 1/5
Focus: Motivation, knowledge of Autodesk, teamwork, adaptability, career goals
A conversational close-out covering what you know about Autodesk (AutoCAD, Fusion, Revit and its Bengaluru/Pune R&D centres), five-year goals, learning adaptability and teamwork scenarios. In the reported drive all four finalists who reached this round received offers.
What they evaluate: Culture fit, genuine interest in Autodesk products, and communication.
Tips: Know 2-3 Autodesk products and what the India engineering centres build; generic answers stand out badly here.
Software Engineer / SDE-1 (Bengaluru / Pune)
Difficulty: Medium · Timeline: ~1 month · Recommended prep: 6-8 weeks · Cooldown: ~6 months (typical, unofficial)
-
Round 0: Online Assessment (CodeSignal)
Assessment · ~70-90 min · difficulty 2/5 · elimination round
Focus: Arrays, strings, hashing, DP, problem decomposition
Autodesk screens SDE-1 applicants with a CodeSignal assessment of 4 coding questions at easy-to-medium (occasionally medium-hard) level, testing algorithms and data structures. String-processing tasks show up often - a reported onsite follow-up was breaking a sentence into dictionary words, so practice that family.
What they evaluate: Raw problem-solving baseline before any interviewer time is spent.
Sample question: Word Break - given a string with no spaces and a dictionary, decide if the string can be segmented into a space-separated sequence of dictionary words. Input: the string, then N, then N words. Output: true or false.
Tips: CodeSignal scores partial progress - keep code compiling and pass what you can on all 4 rather than perfecting 2.
-
Round 1: Technical Interview 1 (DSA + OOP/SOLID)
Onsite · ~60 min · difficulty 2/5 · elimination round
Focus: Trees, binary search, OOPs, SOLID principles, Java/C++ language basics, DBMS
Live coding of 1-2 LeetCode-style problems (binary tree height, left view of a binary tree and binary search with complexity analysis are reported) plus fundamentals. SOLID principles come up explicitly - a Bengaluru candidate was asked to explain all five - along with OOP concepts like runtime polymorphism, ACID properties and language basics for your primary stack (Java for platform teams, C++ for core-product teams).
What they evaluate: Coding fluency on easy-medium DSA plus solid OOP/design vocabulary.
Sample question: Left View of a Binary Tree - print the first node visible at every level when the tree is viewed from the left. Give both the recursive and the queue-based level-order approach with complexities.
Tips: Prepare one-line definitions plus a concrete example for each SOLID principle; reciting definitions alone was not enough in reports.
-
Round 2: Technical Interview 2 (practical coding + basics of scale)
Onsite · ~60 min · difficulty 3/5 · elimination round
Focus: File/CSV parsing, data aggregation, nested-structure manipulation, caching basics, CAP theorem
A practical round closer to day-to-day work: reported tasks include reading a CSV file and running data analysis over it, and flattening a deeply nested array using only loops. For platform/web-leaning roles expect early system questions - caching scenarios when a downstream contract changes, searching large datasets efficiently, handling traffic spikes and a CAP-theorem explanation.
What they evaluate: Ability to write clean working code against messy real-world input, plus first-principles reasoning about caching and scale.
Sample question: Read a CSV of (user, product, amount) purchase rows and print the top 3 users by total spend, handling malformed rows gracefully. Then: how would you change your design if the file were 100 GB?
Tips: Practice quick file-parsing katas in your main language; candidates who only grind LeetCode report being surprised by this round.
-
Round 3: Hiring Manager Round
Behavioral · ~45 min · difficulty 1/5
Focus: Projects deep-dive, motivation, team fit, ways of working
A conversational round with the hiring manager - reported as friendly with no trick questions. Expect a deep dive into resume projects (what you built, challenges, decisions), why Autodesk, and how you collaborate. Recent Pune interviews also asked how candidates use AI tools in their workflow.
What they evaluate: Ownership of past work, communication, and fit with the team in Pune or Bengaluru.
Tips: Have an honest, specific answer on how you use AI coding tools - it is an increasingly common question at Autodesk India.
Software Engineer 2 / SDE-2 (Bengaluru / Pune)
Difficulty: Medium · Timeline: ~3-5 weeks · Recommended prep: 6-8 weeks · Cooldown: ~6 months (typical, unofficial)
-
Round 0: Online Assessment (medium-hard DSA)
Assessment · ~90 min · difficulty 3/5 · elimination round
Focus: Heaps, hashing, graphs, DP - medium and hard LeetCode-style problems
The SDE-2 pipeline opens with an online assessment containing medium and hard DSA questions (reported on the SDE-2 loop via enginebogie). Same CodeSignal-style format as SDE-1 but with the bar shifted up a notch - expect at least one problem needing a heap, graph traversal or non-trivial DP.
What they evaluate: Whether your DSA depth matches the mid-level bar before the panel invests four rounds.
Sample question: Top K Frequent Elements - given N integers and K, print the K values with the highest frequency (break frequency ties by smaller value first). Input: N K, then N integers.
Tips: Revise heap/bucket variants of frequency problems - they are a favourite at this level and easy to fumble under a timer.
-
Round 1: Technical Interview 1 (medium DSA + design discussion)
Onsite · ~60 min · difficulty 3/5 · elimination round
Focus: Graphs/topological sort, trees, medium DSA, HLD warm-up, project discussion
Roughly 15-20 minutes on your current project and decisions, then a medium DSA problem, then a short high-level-design discussion (the reported SDE-2 loop mixes medium DSA with HLD in the same round). Bengaluru candidates also report follow-ups on migration work - e.g. moving services from Node.js to Spring Boot - and CI/CD practices.
What they evaluate: Mid-level coding fluency plus early signal on architectural thinking and real project ownership.
Sample question: Course Schedule - given N tasks and M prerequisite pairs, determine whether all tasks can be completed, and output one valid order if so. Discuss cycle detection via DFS colours vs Kahn's algorithm.
Tips: Rehearse a 3-minute crisp story of your most complex project - this round consistently starts there and the coding clock only starts after.
-
Round 2: Technical Interview 2 (hard DSA + domain)
Onsite · ~60 min · difficulty 4/5 · elimination round
Focus: Hard DSA, computational geometry flavour (core products) or Java/Spring/AWS internals (platform)
The reported SDE-2 loop closes its coding bar with a hard DSA problem plus domain questions matched to the team. Core-product teams (Fusion, AutoCAD, geometry engines in Pune/Bengaluru) lean C++ - expect memory management, STL internals and geometry-flavoured problems. Platform/cloud teams probe Java, Spring Boot, microservices and AWS services you have actually used.
What they evaluate: Depth: can you handle a genuinely hard problem and speak credibly about your domain stack.
Sample question: Given N axis-aligned rectangles (x1 y1 x2 y2 each), compute the total area covered by their union (overlaps counted once) - a natural fit for a CAD company. Input: N, then N rectangles. Output: the area.
Tips: Ask which team the panel is from - C++/geometry vs Java/AWS preparation diverges sharply at Autodesk and the recruiter will tell you.
-
Round 3: System Design (HLD on AWS)
System_design · ~60 min · difficulty 4/5
Focus: API design, SQL vs NoSQL modelling, caching, queues, AWS building blocks (EC2, S3, RDS, SQS)
A dedicated high-level design round in the "how would you design and scale X" / "build X on AWS - which components and why" style that Autodesk interviewers favour. Candidates report caching-invalidation scenarios, searching very large datasets, and absorbing traffic spikes. Grounding choices in AWS services (S3 for artifacts, RDS vs DynamoDB, SQS for decoupling) lands well because Autodesk's platform runs on AWS.
What they evaluate: Structured design thinking: requirements, API, data model, scaling story and honest trade-offs.
Sample question: Design a cloud storage and versioning service for large CAD design files (think Autodesk Docs): upload of multi-GB files, version history, sharing across a team in Pune and a client in the US, and fast preview generation. Which AWS components would you use and why?
Tips: Always state SQL-vs-NoSQL and consistency trade-offs explicitly - CAP theorem questions recur in Autodesk India reports.
-
Round 4: Hiring Manager (behavioral + light LLD)
Behavioral · ~45-60 min · difficulty 2/5
Focus: Past projects, behavioral, ownership, light LLD/HLD probes, team fit
The manager round mixes behavioral questions with a walk through past projects and, per the reported SDE-2 loop, light low-level and high-level design probes - class modelling for a feature you built, or how you would re-architect a component you own. Conversational in tone; managers in Pune and Bengaluru reportedly focus on fit rather than trick questions.
What they evaluate: Ownership, collaboration, design judgement at the boundary of your past work.
Tips: Bring one story where you pushed back on a design decision with data - mid-level hires are expected to show that muscle.
Senior Software Engineer (Bengaluru / Pune)
Difficulty: Medium · Timeline: ~3 weeks · Recommended prep: 4-6 weeks · Cooldown: ~6 months (unofficial)
-
Round 0: Recruiter Screen
Phone · ~30 min
Focus: Background, role fit, compensation expectations, process overview
A short, straightforward chat covering your background, salary expectations and role fit - a November 2025 Bengaluru senior hire (8 YOE) described it as smooth. The recruiter clarifies whether the team is core-product (C++/geometry) or platform (Java/React/AWS), which determines what the panel will probe. Whole process took ~3 weeks for that candidate.
What they evaluate: Basic alignment on experience, expectations and team match.
Tips: Ask directly which product group and stack the panel is from - it changes your prep more at Autodesk than at most companies.
-
Round 1: Technical Coding (experience deep-dive + problem)
Onsite · ~60 min · difficulty 3/5 · elimination round
Focus: Project deep-dive, medium DSA/design-coding hybrid, tree/recursion modelling, clean code
Reported to open with ~20 minutes on your current role - what you built, challenges, decisions - before moving to a coding problem. The November 2025 Bengaluru senior loop featured a "filesystem with boxes-within-boxes" design-and-code question (medium): model a nested container hierarchy and answer aggregate queries over it. Less about exotic algorithms, more about clean modelling of a recursive structure.
What they evaluate: Whether 5-10 years of experience shows up as clean abstractions and calm, communicative coding.
Sample question: Design a filesystem of boxes within boxes: each folder contains files (with sizes) and sub-folders to arbitrary depth. Implement total-size-of-folder and find-largest-folder, and discuss how you would handle moves and cycles.
Tips: Narrate trade-offs as you code (recursion vs iteration, where you would cache sizes) - the deep-dive and the problem are scored together.
-
Round 2: System Design (HLD, AWS-grounded)
System_design · ~60 min · difficulty 4/5 · elimination round
Focus: Scalable architecture, API and data modelling, AWS components, microservices, resilience
The classic Autodesk design round: "how would you design and scale X" and "build X on AWS - which components would you use and why". For seniors, system design plus cloud (AWS) experience is effectively mandatory; panels may include up to 3 interviewers. Web/platform candidates should be fluent in microservices on AWS (EC2/EKS, S3, RDS, SQS/SNS); core-product candidates get an equivalent architecture discussion around their C++ services.
What they evaluate: Ability to own an ambiguous design end-to-end: requirements, APIs, data, scaling, failure modes, cost.
Sample question: Design a collaborative viewer for large 3D design files: engineers in Pune, Bengaluru and San Francisco open multi-GB models in a browser, see teammates' live annotations, and work through flaky networks. Cover derivative/preview generation, CDN strategy and conflict handling.
Tips: Anchor every choice to an AWS service and a reason; vague "we will use a cache" answers were the reported differentiator between loops.
-
Round 3: Principal Engineer Round (resume review + HLD)
Onsite · ~90 min · difficulty 4/5 · elimination round
Focus: Career-depth resume review, second HLD problem, technical judgement, depth probes
A distinctive ~90-minute round with a principal engineer: a significant portion is a detailed resume review - expect probing on why decisions were made across your career - followed by a high-level design question (a URL-shortener-class problem was reported). This is where the senior bar is really set; superficial answers about past architecture get exposed.
What they evaluate: Technical depth and honesty over your whole track record; design judgement under senior scrutiny.
Sample question: Design a URL shortener at global scale, then extend it: custom aliases, expiring links, per-tenant analytics, and how your design changes if reads are 1000x writes. Be ready to defend ID-generation and storage choices in depth.
Tips: Re-read your own resume critically before this round - every line is fair game, including projects from years ago.
-
Round 4: Hiring Manager / Behavioral
Behavioral · ~45 min · difficulty 1/5
Focus: Leadership stories, mentoring, conflict, ways of working, team fit
A conversational wrap-up with the hiring manager focused on fit: mentoring juniors, handling disagreement, working across time zones with US teams, and why Autodesk. Described in reports as friendly with no tricky questions - but seniors are expected to bring concrete leadership stories, not just technical ones.
What they evaluate: Senior-level behaviours: mentorship, influence without authority, and communication.
Tips: Prepare STAR stories on mentoring and on a project you steered back on track - both themes recur in Autodesk senior loops.
Other companies' interview processes
Worked sample solutions (code, design diagrams, and STAR answers) for every round are available in the FAANGPlus app: open interview processes.