Uber's SDE-2 loop in India is reported as an online assessment and a screen, then several further rounds mixing DSA, design and a hiring-manager conversation. One candidate called it very difficult, and an SDE-2 author says the low-level design round expects working code.
8 rounds
2-6 weeks (one guide), 4-6 weeks (another); one candidate report says about 5 weeks
Reported by candidates and, where we found one, the company itself. Each round shows how well it is reported. Processes change by team, level and year.
1
Online assessment
Online assessment70-90 min (one guide)
Several reports
Timed coding test, reported on CodeSignal by an SDE-II candidate and by one guide. A 2025 intern candidate described three problems based on real-world challenges.
algorithmic coding problems
speed under a timer
Varies: A Bengaluru SDE-II report (Feb 2025) and a 2025 intern report both have it. A senior Bengaluru report (Oct 2024) starts with a hiring-manager interview and does not mention one.
What can be asked in a online assessment round
A timed test on a coding platform, taken alone. Usually two or three coding problems, sometimes with multiple-choice questions on CS basics, aptitude or SQL.
Array and string problems that need a hash map, two pointers or a sliding window
Read the constraints first, because they point to the technique. If n is 100,000, an O(n^2) scan will time out, so look for a hash map, two pointers on a sorted array, or a window that moves forward only. Sketch the brute force, find the repeated work, then remove it. For two-sum style pairs you might say: I store each value's index as I scan, so the complement lookup is O(1) and the whole pass is O(n).
Greedy and sorting problems with a catch in the constraints
Greedy problems hide behind a sort. Ask what ordering makes the locally best choice safe, then try to argue why swapping two adjacent choices never helps. Interval scheduling is the classic case: sort by end time and take every interval that starts after the last chosen one. Check the catch in the constraints, such as overflow, ties or equal values, and test a tiny case by hand before submitting.
A medium dynamic programming or graph-on-a-grid problem
Decide whether the problem is a graph walk or a recurrence. On a grid, a shortest path with equal step cost is breadth-first search from the start, with a visited set so each cell is processed once, giving O(rows x cols). For dynamic programming, define the state in one sentence, write the transition, set the base cases, then see if only the previous row is needed to save space.
Start from the table that holds the thing being counted, join the others on their keys, filter rows with WHERE, group, then filter groups with HAVING. Remember that WHERE runs before grouping and HAVING after. Use LEFT JOIN when rows with no match must still appear, and COUNT of a column rather than star when nulls matter. Example: customers with more than three orders is GROUP BY customer_id HAVING COUNT(*) > 3.
Multiple-choice questions on OOP, operating systems, databases and networks
These reward quick, accurate recall rather than long reasoning, so revise the short list of definitions that recur. Know the difference between a process and a thread, the four ACID properties, what an index trades for faster reads, TCP against UDP, and the four pillars of OOP. Eliminate two wrong options first, then choose. Do not spend more than a minute on any single question, and flag it to return to if time remains.
Recruiter or phone screenabout 40-60 min (guide estimate)
Several reports
A live call. The Feb 2025 SDE-II report describes some general computer-science questions followed by one DSA question. One guide says the engineering manager of the hiring team runs it, with a coding problem and a short work-experience chat.
core CS basics
data structures and algorithms
work experience (one guide)
Varies: Who runs it and which platform (CodeSignal or HackerRank) differ between sources.
What can be asked in a recruiter or phone screen round
A short call that decides whether you move on. Part resume and motivation, sometimes one easy coding or fundamentals question in a shared editor.
Walk me through your resume and your best project
Give a two minute spine, not a life story: who you are now, the one or two projects that matter most, and why this role is the next step. Pick the best project and state the problem, your part, the technology and one measurable result. Keep every claim something you can defend if the next question digs in. You could open with: I am a backend engineer, and the project I am proudest of is the one where I cut checkout latency.
The listener wants to know whether your reason is specific and sincere, not flattery. Name one or two real things about the product, the engineering problems or the team that you have looked at, then connect them to something you have done or want to learn. Why now should be about your growth, not escaping your current job. An opening could be: I have used the product, and the scaling problem it describes is the kind I want to work on.
Notice period, location and compensation expectations
Give the facts that are true and keep the negotiation for later. State your notice period exactly as your contract says, and mention whether it can be shortened or bought out only if that is real. For compensation, say you would like to understand the role and level first, or give a researched range rather than a single anchor. Never invent a current figure. If you are asked directly, answer honestly and calmly.
One easy or medium coding question, or a language fundamentals question
Treat it as a short version of the full coding round. Repeat the problem back, ask about input size and edge cases, state a plan before typing, and talk while you code. For a language question, such as what a closure is or how a hash map handles collisions, give a one line definition and a tiny example. A good line is: first I will do the simple approach so we have something correct, then improve it.
Data structures and algorithmsabout 60 min each (guide)
Several reports
Two DSA-heavy rounds in the Feb 2025 SDE-II loop, the second blended with coding or design. One candidate was asked tree and graph problems (binary tree diameter, distance between two graph points). A guide calls the problems mostly medium difficulty and expects code that compiles and runs.
trees and graphs
data structures and algorithms
code that runs
Varies: Uber's own careers story for a backend engineer lists three coding interviews; the 2025 SDE-II report lists two DSA-heavy rounds.
What can be asked in a data structures and algorithms round
A live coding round of 45 to 60 minutes with one or two problems. You are judged on how you clarify, reason, code, test and state the complexity, not on the answer alone.
Arrays, strings and hashing: pairs, windows, frequency counts, prefix sums
Name the pattern from the shape of the question. Pairs with a target use a hash map of seen values. A contiguous range with a condition uses a sliding window that grows on the right and shrinks on the left, so each element enters and leaves once, giving O(n). Range sums over many queries use a prefix sum array so each query is O(1) after O(n) setup. Count characters with a fixed-size array or map.
Linked lists, stacks and queues, including design-a-structure problems such as an LRU cache
For an LRU cache, combine a hash map from key to node with a doubly linked list ordered by recency. Get moves the node to the front, put inserts at the front, and when over capacity you remove the tail node and its map entry. Both operations are O(1), and you say why: the map finds the node, the list reorders it without a scan. For stacks, mention a monotonic stack for next greater element.
Trees and BSTs, graphs (BFS, DFS, topological sort, shortest paths) and grids
Decide the traversal from what is asked. Level by level or shortest path in an unweighted graph is BFS with a queue. Exploring all paths, components or cycles is DFS. A dependency order is topological sort: Kahn's algorithm repeatedly removes nodes with in-degree zero, and leftover nodes mean a cycle. Weighted shortest path with non-negative edges is Dijkstra with a heap, O((V + E) log V). On a BST, an in-order walk is sorted.
Heaps and top-K, intervals, binary search on the answer
For the K largest items keep a min-heap of size K: push each element and pop the smallest when the heap exceeds K, so the cost is O(n log K) and memory is O(K). For intervals, sort by start, then merge by comparing the next start with the current end. For binary search on the answer, define a yes or no check that is monotonic, such as can ship within D days, and search the capacity range.
Dynamic programming from one dimension up to knapsack and string DP
Define the state in a sentence, write the recurrence, fix the base cases, and decide the order of filling. Start with the recursive version with memoisation, then convert to a table. For 0/1 knapsack, dp[w] is the best value at capacity w, and in the one dimensional form you loop capacity downward so each item is used once. For string problems such as edit distance, the state is a pair of prefix lengths.
Follow-ups: optimise space, handle streaming input, make it thread-safe
Treat follow-ups as a chance to show depth, and answer in the same calm order each time. To save space, ask which previous rows or values the recurrence actually reads and keep only those. For streaming input, say what summary you can keep, such as a heap of size K or a running count. For thread safety, name the shared state, guard it with a lock, and keep the critical section short.
An object-oriented design problem. One SDE-2 author says LLD requires working code, and a guide describes a machine-coding round for 2026 SDE-2 candidates that emphasises OOP and extensibility. A 2025 intern loop listed an LLD round; the Feb 2025 SDE-II report has no separate one.
object-oriented design
working extensible code
Varies: Not every report lists a separate LLD round; the Feb 2025 SDE-II loop has a 'DSA plus coding/design' round instead. A 2025 intern loop had a separate LLD round.
An object-oriented design discussion: classes, interfaces, relationships and the patterns that keep them flexible. Often combined with machine coding.
Class design for a booking system, a notification service or a payment module
Begin by listing the nouns and verbs from the requirements, then turn nouns into classes and group verbs by who owns the data. For a booking system you might have Resource, Booking, User and a BookingService, with a rule that two bookings for one resource cannot overlap. Draw relationships clearly: composition when a part cannot exist alone, an interface when behaviour varies. Say which classes you would write first and why.
Tie each principle to a concrete choice in your design rather than reciting the acronym. Single responsibility: the invoice class does not also send email. Open and closed: add a new payment method by adding a class, not editing a switch. Liskov: a subclass must honour the parent's contract. Interface segregation: small interfaces. Dependency inversion: the service depends on a notifier interface, not on an SMS client.
Strategy, observer, factory, state and builder patterns, and when each helps
Know the problem each pattern solves. Strategy swaps an algorithm, such as pricing or routing. Observer lets many listeners react to one event, as in notifications. Factory hides which concrete class is created. State moves behaviour into a class per state, as in an order lifecycle. Builder assembles an object with many optional fields. Say when not to use one: a pattern with a single implementation is usually noise.
Concurrency inside the design: locks, thread-safe collections, idempotent operations
Identify what is shared and mutable, then choose the smallest protection that works. Use a lock around a compound check-then-act, a concurrent map for independent keys, and atomic counters for simple tallies. For operations that may be retried, make them idempotent by carrying a request id and ignoring a repeat. Example: charging a booking twice is prevented by storing the request id with the result and returning it again.
Extending the design for a new requirement without breaking old code
Show the design absorbing change. When a new requirement arrives, say which existing abstraction already covers it and which new class you add. If a discount rule appears, add a new implementation of the discount interface and register it. If you must change old code, explain why the original split was wrong and how you would limit the edit. Mention tests: existing tests should still pass unchanged after the extension.
A design discussion. One candidate was asked to walk through their most significant past project; another reported being asked to design a car-rental system and a stock-ticker system. Uber's own careers story for a backend engineer lists two design and architecture interviews.
system design
explaining a past project
Varies: One guide says mid-level and senior candidates get one or two rounds. The reported prompts differ between candidates.
A 45 to 60 minute discussion of how to build a system at scale: requirements, numbers, APIs, data model, components, bottlenecks and trade-offs.
Short links, chat, news feed, notifications, search, rate limiting
Use one repeatable path: clarify scope and scale, estimate load, define the API, sketch the data model, draw the main components, then probe bottlenecks. For short links, store a mapping from a generated short id to the long url, generate ids with a counter encoded in base 62, cache hot links, and note reads far outnumber writes. For chat, mention persistent connections and a store ordered per conversation.
Payments, wallets and ledgers where money must never be lost or doubled
Money systems put correctness before speed. Use a double-entry ledger where entries are append-only and every transfer writes a debit and an equal credit in one transaction. Protect against doubled charges with an idempotency key stored with the result, so a retry returns the original outcome. For lost updates use a database transaction, and reconcile with the payment provider's records on a schedule.
Ride matching, maps and anything with live location
Split the problem into location ingestion, a geospatial index and matching. Drivers send updates every few seconds to a write-optimised store, often in memory, keyed by a grid cell such as a geohash or S2 cell. To find nearby drivers, query the rider's cell and its neighbours, then rank by distance and estimated time. Say how you handle a stale location and a driver who accepts two rides.
File storage and sync, video streaming and content delivery
For file sync, split files into chunks, hash each chunk and upload only chunks the server lacks, so a small edit transfers a small amount. Keep metadata in a database and chunks in object storage. For video, upload the original, transcode into several resolutions, store segments and serve them through a content delivery network with adaptive bitrate streaming, so the player switches quality as bandwidth changes.
Treat each as a tool with a cost. A cache cuts read latency but adds staleness and invalidation. Sharding spreads write load but makes cross-shard queries hard. Replication gives availability and read scale but introduces lag. Queues smooth spikes and decouple services but require idempotent consumers. For failures, name timeouts, retries with backoff and a fallback. State what you give up whenever you choose.
Estimates are there to choose a design, so finish each one with a decision. Use round numbers: a day has about 100,000 seconds, so 10 million requests a day is roughly 100 per second on average, with a peak of a few times that. Multiply object size by count for storage, and compare against one machine's capacity. If one database handles the load, say so and avoid sharding.
A conversation with the hiring manager, listed as the last round in the Feb 2025 SDE-II report. One guide describes a collaboration and leadership round on past projects or how you would deliver future ones, and one candidate advised not to underestimate it because it carried more weight.
past projects
collaboration
leadership
handling failure (one guide)
Varies: A senior Bengaluru report (Oct 2024) says the hiring-manager interview came first, followed by several parallel rounds; the Feb 2025 SDE-II report places it last. Uber's careers story names a collaboration and leadership interview but not a hiring manager.
A conversation about your past work and how you would fit the team: depth on projects, ownership, decisions and trade-offs, sometimes a design problem.
Pick a project and go deep: why this design, what broke, what you would change
Choose a project you can discuss at every layer. Explain the problem and who it served, the options you considered, why you chose one, what broke and what you would change now. Be clear which parts you did and which the team did. Admitting a flaw earns trust. Opening line: the main constraint was latency, so I chose a cache in front of the database, and I would now add an invalidation strategy.
A time you disagreed with a teammate or pushed back on a decision
Use situation, task, action, result, and keep the disagreement about the work, never the person. The listener is checking that you can hold a view with evidence, listen, and commit once the decision is made. Describe what you proposed, the data or reasoning you used, how you reached agreement or accepted the other path, and what happened. Do not invent a story: use a real one, even a small one.
Show a method rather than a claim of being busy. Describe how you judge impact and deadline, ask who needs the result and when, and say no or renegotiate scope early when two things clash. Mention communicating the trade-off to stakeholders so nobody is surprised. A sentence you could adapt: I list what is blocking others first, then what has a fixed date, then what moves a metric.
Answer honestly and tie it to what the role offers. Say what you want to learn or own next, such as larger systems, more ownership of design, or mentoring, and what kind of team helps you do it. Then ask what the team is trying to achieve this year and how the work is reviewed. The listener checks that your wishes match the job and that you have thought about it.
One guide describes an interviewer from outside the hiring team running a deep dive into one past project, used to help decide your level. Only that guide mentions this round, and none of the candidate reports read here list it.
depth on a past project
leveling signal
Varies: The guide says staff candidates need a project spanning at least two to three quarters, while senior candidates can use a less complex one.
An independent interviewer from outside the hiring team who judges you against the company's leadership principles, with the power to hold the bar.
Stories mapped to leadership principles such as ownership and customer obsession
Prepare a handful of real stories and be ready to map each to more than one principle. Choose the story that fits the principle being probed. Use situation, task, action and result, speak in the first person about what you did, and include a measurable result where you have one. The listener checks that your behaviour matches the principle, not that you know its name.
Deep follow-ups on every claim in a story: what exactly you did, what the numbers were
Expect repeated why and how questions on one story, to separate your own work from the team's. Know the details: the dates, the numbers, the alternatives you rejected, and who disagreed. Say we only when it is true, and I when it is you. If you do not remember a figure, say so and give your best estimate. Consistent detail is persuasive; invented detail collapses under the next question.
A technical or design question to confirm the level
This is a shorter check that your technical level is real. Treat it like any design or coding question: clarify, state assumptions, give a simple solution first, then discuss trade-offs and failure cases. The listener looks for judgement, such as knowing what to leave out and when to be careful. Explaining your reasoning aloud matters more than a perfect final design.
One guide says results are reviewed by a hiring committee after the loop, and that most candidates hear back within three to five business days. Another guide describes the interview panel meeting to decide, with team matching either at the start or the end.
overall loop signal
level fit
Varies: One guide says matching can happen at the start or the end and can take weeks to months. A 2023 Hyderabad campus drive had four rounds and ended with a managerial and HR interview instead.
The closing conversation: motivation, expectations, notice period, compensation and how the offer is structured. Prepared answers and a clear number help here.
Current and expected compensation, and how to answer without boxing yourself in
Be honest and avoid giving a single number too early. Share your current pay only if you are comfortable, and make clear what is fixed and what is variable. For expectations, say you are flexible for the right role, and offer a researched range based on level and market. If pushed, state a figure you can justify. Never state a number you cannot support, because offers are often checked.
Give the facts exactly as they are in your contract. State the notice period, whether early release or a buy-out is truly possible, and the earliest honest joining date. Do not promise a date your employer has not agreed to. For relocation, say whether you can move, how soon, and whether you would need support. Clear, early information helps the recruiter plan and shows you are reliable.
Keep it positive and forward looking. Talk about what you want more of, such as larger scale, ownership or a stronger team, rather than criticising your current employer. The listener checks that you will not leave this job for the same reason. Opening line: I have learned a lot in the current role, and I now want to work on problems at a bigger scale.
Levels, equity and how the offer is built
Ask for the structure rather than only the total. Find out the level and title, base salary, bonus and whether it is guaranteed, equity amount, vesting schedule and cliff, joining bonus and any clawback. Compare offers on a yearly total with the same assumptions. If equity is in a private company, ask how it is valued and what happens on leaving. Asking calmly is normal and expected.
The hiring manager, behavioural and offer conversations. Candidate reports shape the themes; the answers and frames are written for this site, so use them as a structure and bring your own stories.
We build globally, we live locallyUse worldwide scale while staying close to local cities, communities, drivers and riders.
We are customer obsessedEarn customers' trust by helping them save money or earn more, and accept short-term costs for long-term loyalty.
We celebrate differencesMake people from all backgrounds welcome and invite varied views before the team commits to a plan.
We do the right thing. Period.Act ethically without qualification, whatever the pressure.
We act like ownersTake initiative on problems, support colleagues, act decisively and take responsibility for errors.
We persevereTake on hard challenges rather than easy paths, and treat resilience as a core strength.
We value ideas over hierarchyGood ideas can come from anyone; test them through open debate and then act.
We make big bold betsAccept that some bets fail, learn from them, and keep moving forward.
Themes candidates report being probed
Working on a team
Dealing with failure
Leadership
Past project delivery and impact
Domain expertise and confidence
Ownership with numbers
Questions to rehearse
Walk me through the most significant project you have worked on.Several reports
What is being checked. A Bengaluru SDE-2 candidate lists a walkthrough of their most significant project, and a guide says the hiring manager walks through past projects. The listener checks depth, ownership and the scope you handled for your level.
How to frame it. Prepare a five-minute walkthrough: the problem, your role, the architecture in broad strokes, two trade-offs, and the result with a number. Be ready to go deeper on any piece. Opening to adapt: "I led [component] of [project], which handled [scale], and the key decision was [choice]."
How did you deliver a past project, where did it go wrong, and what was the measured impact?Prep guides
What is being checked. A guide lists delivery, failure and quantified impact together. The listener checks honesty about problems and whether you track results with numbers.
How to frame it. Give a timeline with milestones, one setback and how you recovered, and the final metric (latency, cost, adoption). Admit what slipped. Opening to adapt: "We planned [milestones], slipped on [one], and ended up with [result] measured by [metric]."
Pick one past project and go into its technical complexity, trade-offs and what failed.Prep guides
What is being checked. One guide describes a Bar Raiser deep dive on a single project, and says the project is a level input. The listener checks that the project's scope matches the level you are targeting.
How to frame it. Choose a project at or slightly above your target level. Cover the complexity, alternatives you rejected and why, time to milestones, and one failure. Keep claims separate from your team's. Opening to adapt: "The hardest part was [problem], and I weighed [A] against [B] because [reason]."
How would you describe your domain depth, and how have you worked with and guided a team?One report
What is being checked. A senior Bengaluru candidate reports the main hiring-manager conversation covering domain expertise, team aspects, management skills and confidence. The listener checks leadership signal.
How to frame it. State your core domain in one sentence, then give two examples of guiding others (a design review you led, a person you grew) with outcomes. Speak with calm confidence. Opening to adapt: "My strongest area is [domain], where I have [concrete achievement], and I lead by [method]."
Tell me about pushing through a difficult problem when it would have been easier to stop.Prep guides
What is being checked. A guide stresses ownership, past impact and numbers. This also ties to the Persevere and Act like owners norms. The listener checks grit and sensible judgement about when to push.
How to frame it. Describe the obstacle, what you tried that failed, what finally worked, and the measurable result. Show that you asked for help when it was sensible. Opening to adapt: "For weeks [problem] resisted every fix I tried, until I [turning point]."
Tell me about a time your idea was challenged by someone more senior, and what happened.Prep guides
What is being checked. Ties to the norm of valuing ideas over hierarchy. The listener checks that you argue on merit, listen, and commit after a decision.
How to frame it. Explain your idea and the senior person's concern, the evidence you each brought, and the outcome. If you lost, show that you committed and helped it succeed. Opening to adapt: "[Senior person] preferred [approach], so I showed them [data], and we agreed to [decision]."
Tell me about a time doing the right thing cost you time or convenience.Prep guides
What is being checked. Ties to the do-the-right-thing norm. The listener checks integrity under pressure and whether you raise concerns rather than stay silent.
How to frame it. Describe the shortcut that was available, why you did not take it, who you told, and what it cost. Keep it factual and avoid moralising. Opening to adapt: "We could have shipped [shortcut], but it risked [harm], so I raised it with [person]."
A guide reports that the recruiter call covers salary, location and level expectations, and advises against revealing your salary expectations or the state of other processes.
The same guide says the panel decides after the loop, that one strong no-hire can end the process, and that the Bar Raiser project is a leveling input.
A separate guide reports that most candidates hear back within three to five business days of the loop.
General advice: ask which level is being offered, get the full breakdown in writing, and check your notice period and buy-out before agreeing a start date.
What we could not verify. The values source is a November 2017 post that carries an October 2021 update note pointing to newer values that could not be retrieved, and one guide names different values (Trip Obsessed, Go Get It, Build with Heart), so the list here is likely out of date. The Bar Raiser round comes from one guide that neither first-hand snippet confirms, and Uber's careers site gave no interview description.
What we could not verify. Uber's own careers pages are undated and give no durations or India-specific rounds. LeetCode, Glassdoor and 1point3acres threads (the richest recent SDE-2 reports) returned 403, so round order and the 2025-26 LLD round rest on a few Jointaro/Aced pages plus guides. Bar raiser and committee are single-guide claims. Duration figures mostly come from guides, and the 2026 LLD detail is one guide. The Feb 2025 report gives no durations.