Uber runs a highly standardised, named-round loop that candidates describe almost identically across geographies: a timed online assessment (CodeSignal, HackerRank or Codility), a screening or BPS coding call that acts as the elimination gate, then a same-day set of one-hour rounds with fixed labels - Algorithms & Data Structures, Depth in Specialization (production-quality machine coding), Design & Architecture on a new problem, Design & Architecture on a previously solved problem, and a Collaboration/Leadership or bar-raiser conversation with a manager. Recruiters brief candidates on the round names and send prep material in advance, so the loop is unusually predictable; the bar inside it is not, with several writers reporting rejections for readability, missed edge cases or a project story judged not to be at Uber scale. Non-engineering loops diverge sharply: product management runs a multi-hour take-home plus a group-style product case day with PM, engineer and UX-designer interviewers, and analytics roles pair SQL/Python screening with city-launch and marketplace case studies.
These 22 writeups cover software engineering, data & science, product management and other roles, India-heavy, some US and Europe. 17 of them were posted under a pseudonym (Reddit, GeeksforGeeks or forum handles), so we could not verify the authors’ identities.
Typical rounds
4.32-7 range
Outcomes shared
8/7offer / not
Most common round
Technical
Sources span
2017 - 2025
Most frequently reported · Data structures & algorithms(18), Behavioral(17), Past projects(17), Low-level design(10), Dynamic programming(9)
Uber Technologies’s official process
Uber's own "How we hire" page lays out a seven-step pipeline: apply, a talent-team call, a hiring-manager conversation, a technical interview for technical roles only, an optional role-dependent functional exercise (analytics task, written exercise, portfolio review or work simulation), a team interview with cross-functional partners where any exercise is presented back, and a recruiter-led debrief and decision. It stresses that every role differs and points candidates to per-team pages. Separately, Uber commissioned a public candidate-prep site for software roles (https://s3.amazonaws.com/ubercandidateprep/index.html, built with Gayle Laakmann McDowell) covering the process overview, behavioural questions, design/architecture questions, Big-O and a seven-step algorithm method, which matches the round names candidates report.
A single medium-difficulty coding problem acting as the elimination gate; the author notes clearing it books all four remaining rounds regardless of later performance. Class structure and naming are graded alongside correctness.
Covered · Data structures & algorithms, OOP
2
Algorithms and Data StructuresTechnical
Roughly fifty minutes of problem solving after introductions. The author worked through a DP problem, a follow-up on it, and a third problem answered verbally without implementation.
Covered · Dynamic programming, Data structures & algorithms
3
Depth In SpecializationLow-level design
A second coding round graded on production-readiness rather than algorithmic cleverness: single-responsibility classes, sensible method and variable names, readable structure.
Covered · Low-level design, OOP, Testing
4
Design and Architecture (New Problem)System design
A greenfield design brief where the candidate clarifies requirements, estimates capacity, gives a high-level architecture and then drills into API and schema detail before discussing bottlenecks and scaling.
Covered · System design, Scalability, API design, DBMS
5
Collaboration and Leadership + Previously Solved ProblemHiring manager
A manager-led seventy-five-minute session split between a deep replay of the candidate's own past architecture decisions and their hindsight on them, and questions on leading, resolving conflict and setting team norms.
Covered · Past projects, Behavioral, Conflict handling
A one-hour algorithmic call at roughly medium difficulty, with the author stressing that talking through trade-offs before coding mattered as much as the final solution.
Covered · Data structures & algorithms
2
Coding - Algorithms & Data structuresTechnical
Another medium-level puzzle, run in CodeSignal so the candidate can compile, run and debug during the interview.
An architecture conversation about one of the candidate's own shipped projects: what problem it solved, why each decision was made, how it behaved in production.
Covered · Past projects, System design
4
Coding - Depth in specializationLow-level design
A build-a-small-real-thing round rather than a puzzle, for example a class or library that aggregates statistics over a request stream.
Covered · Low-level design, OOP
5
Design & Architecture - New ProblemSystem design
A classic whiteboard system-design brief on a virtual board.
Covered · System design, Scalability
6
Behavioral - Citizenship and CollaborationBehavioral
Focused on teams and people rather than technology: how the candidate worked with others, what went wrong, what they learned.
Covered · Behavioral, STAR stories, Conflict handling
7
Bar raiserBar raiser
A behavioural interview run by an engineer, drifting back and forth between technical decisions and the people side of them.
Nominally a screen but run as a full coding round with a working solution expected. The candidate discarded a brute-force DFS and a DP idea, then modelled the grid as a 0-1 weighted graph and coded a shortest-path solution inside the hour.
Covered · Graphs, Dynamic programming, Data structures & algorithms
2
Algorithms and Data StructuresTechnical
A single number/string manipulation problem the candidate had seen a variant of before, coded to a working solution within the hour.
Covered · Arrays & strings, Data structures & algorithms
3
Machine codingLow-level design
Implement a small counter class whose entries expire after a fixed window. The candidate offered a priority-queue design, then a timestamp-list plus binary-search design, coded and ran it, but the interviewer wanted stale entries evicted on read and graded it a no-hire. The hiring manager overrode this and let the loop continue.
Covered · Low-level design, Data structures & algorithms, API design
4
System design (bar raiser)System design
Design an internal near-real-time driver heat-map plus a next-day hourly rollup. The candidate wrote requirements and APIs first, rejected in-memory counting at the stated event rate, and proposed geohashing the location stream, windowed aggregation in a stream processor, a cache for the live view and a relational store for hourly aggregates.
Covered · System design, Scalability, Concurrency, API design
5
Tech lead managerHiring manager
A long whiteboard walkthrough of one complex project the candidate had led, with the interviewer pressing on every technical and non-technical obstacle, followed by a pitch of the team's roadmap and motivation questions.
Received an offernot stated in the writeup; author links a separate compensation post6 rounds
1
Online assessmentOnline assessment
A Codility set of four problems spanning easy to hard with a stated pass mark, taken before any human contact.
Covered · Data structures & algorithms
2
Technical phone screenPhone screen
One hard grid problem with no hints offered. The candidate's approach was not the one the interviewer wanted and the feedback was weak, but the recruiter still advanced them with a warning that it would count in the debrief.
Covered · Data structures & algorithms
3
Coding (DSA)Technical
A running-median style problem escalated by a follow-up that loosened the definition of the median and demanded constant time and space; the candidate solved it with bucketed counts.
Covered · Data structures & algorithms, Trees
4
Machine codingLow-level design
A two-word prompt on the shared editor, expanded by the interviewer into a full low-level design with design patterns, three working features and thread-safe concurrency.
Covered · Low-level design, OOP, Concurrency
5
System designSystem design
Design a stock-price indicator, with an extended argument about datastore choice and schema before moving to component trade-offs and scaling.
Covered · System design, DBMS, Scalability
6
Hiring managerHiring manager
Standard manager conversation about current work, past experience and leadership behaviours.
Covered · Behavioral, Past projects, Leadership Principles
An introductory conversation with no technical content, initiated over LinkedIn.
Covered · Resume deep-dive
2
Online assessmentOnline assessment
Two CodeSignal problems at the harder end of medium and above, both of which had to pass every test case to advance.
Covered · Trees, Graphs, Data structures & algorithms
3
Data structuresTechnical
One hard graph problem with variations, again requiring compiling code that passes the interviewer's own test cases. The candidate compared traversal and disjoint-set approaches before implementing one.
Covered · Graphs, Data structures & algorithms
4
Machine codingLow-level design
An extensible object-oriented design implemented end to end in the shared editor, with explicit probing of threading, synchronisation and inter-thread communication.
Covered · Low-level design, OOP, Concurrency
5
System designSystem design
Two halves: first walk through a system the candidate had designed from scratch, then take a fresh design brief with heavy emphasis on handling scale.
Covered · System design, Scalability, Past projects
6
Hiring managerHiring manager
One project explained in depth plus behavioural questions.
A binary-tree question that went well; the onsite was then deferred for months during a hiring pause before the recruiter re-engaged.
Covered · Trees, Data structures & algorithms
2
Algorithms & Data StructuresTechnical
A topological-ordering style problem. The candidate's approach and complexity analysis were right but one edge case was implemented wrongly; feedback flagged readability and the need for hints.
Covered · Graphs, Data structures & algorithms
3
Depth in SpecializationTechnical
A tree traversal problem solved iteratively first, then a follow-up demanding constant extra space that the candidate could not reach in time; feedback said the bar was lowered.
Covered · Trees, Data structures & algorithms
4
Behavioral and design deep-diveBehavioral
An account of the most complex project the candidate had worked on. Feedback said the scope and complexity described were below the level Uber expected.
Covered · Past projects, Behavioral
5
System designSystem design
A chat-service design where the candidate scoped first and compared options; feedback praised the trade-off discussion but wanted more API detail.
Resume walk plus the hardest-project question with follow-ups. Feedback praised communication but questioned leadership depth for the level.
Covered · Resume deep-dive, Behavioral, Past projects
2
CodingTechnical
Design a container supporting constant-time insert, delete, lookup and random removal, followed by edge cases and a multithreading extension. This was the candidate's best-rated round.
Covered · Data structures & algorithms, Concurrency
3
System designSystem design
Design a chat application including delivery-status visibility. The rejection turned on unconvincing reasoning for the datastore choice given the candidate's own capacity estimates.
Covered · System design, DBMS, Scalability
4
Coding on a machineTechnical
A further coding round written and executed on a computer rather than a whiteboard.
A day of back-to-back one-on-one conversations with engineers, rescheduled on the fly by the recruiter. The author recalls a loosely structured day with rooms changing between sessions rather than a fixed round list.
Covered · Past projects, Behavioral
2
Hiring manager lunchHiring manager
A lunch interview where the manager probed willingness to absorb a workload far beyond normal hours; the candidate's qualified answer landed badly.
Covered · Behavioral, Shift availability
3
Remote engineer interviewTechnical
A video call with an engineer from another team, most of it spent on conferencing problems, ending in a discussion of the payments product and what the candidate would change about it.
Covered · Past projects, Product sense
4
Front-end pair interviewTechnical
Two engineers asking front-end fundamentals questions.
Covered · Web development, Data structures & algorithms
5
Recruiter closeFinal
A compensation conversation where the recruiter set the expectation of a below-market base offset by equity. The rejection call came the next day.
Three problems covering base conversion, a binary-string flipping problem and a variation on a well-known interval DP task; the candidate reached the interview shortlist the next day.
Sixty minutes on CodeSignal with three or four problems covering array reconstruction, a maximise-the-minimum cutting task, a bounded grid traversal and a removal-and-merge maximisation. The candidate finished three fully and one partially.
Three problems in an hour on CodeSignal, including base conversion and two competitive-programming style tasks. Sixteen students were shortlisted from the campus drive.
Two interviewers, a binary-search-on-answer variant explained patiently through a test case, then a graph problem.
Covered · Data structures & algorithms, Graphs
3
DesignLow-level design
An object-oriented design brief modelling a retail shop system with added constraints, again with two interviewers.
Covered · Low-level design, OOP
4
HR plus technicalHiring manager
A long closing round covering projects and technology choices, internship conflicts, failure and disagreement stories, strengths and weaknesses, and about fifteen minutes on a logic puzzle.
Covered · Behavioral, Past projects, Puzzles, Conflict handling
Ninety minutes with three problems, two adapted from a competitive-programming archive and one on strongly connected components.
Covered · Graphs, Data structures & algorithms
2
Hiring managerHiring manager
A project discussion centred on failure: which projects went wrong, what the candidate does when stuck, and where motivation was hardest to sustain.
Covered · Past projects, Behavioral
3
TechnicalTechnical
Given an array of node structs, determine whether they form exactly one binary tree containing all of them, coded in full.
Covered · Trees, Data structures & algorithms
4
Technical designLow-level design
Implement predictive-text contact search, discussing which maps and preprocessing are needed and how a trie changes the design, with the full solution coded up.
A graph problem on Uber's own coding platform with live test cases; a dry run resolved a disagreement about the expected output and the candidate was then asked to tighten redundant loops.
Covered · Graphs, Debugging
2
Low-level designLow-level design
Design and fully code a home-device control application with shared and device-specific commands. The prompt was deliberately underspecified, so requirement-gathering questions were part of the assessment.
Covered · Low-level design, OOP, API design
3
HRBehavioral
Introduction, deep questions on the candidate's thesis project and architecture decisions, then behavioural scenarios on teammates who under-deliver and disagreements over technical decisions. The interviewer also drew the candidate into a discussion of Uber's own architecture.
Covered · Past projects, Behavioral, Conflict handling
A meeting-scheduling interval problem, coded, then a variant asking for slots where at least one participant is free, which the candidate could reason about but not code in time.
Covered · Data structures & algorithms, Arrays & strings
3
Low-level designLow-level design
Design a structure supporting insertion, update and related operations, comparing arrays, linked lists, search trees and priority queues, then a deep dive into heap variants and their complexities.
Covered · Low-level design, Data structures & algorithms, OOP
4
Hiring managerHiring manager
Core CS fundamentals applied to real systems, including writing semaphore code, plus past internship stories and questions about ethics and team communication.
Three DSA problems in a campus drive, including an interval DP task and a weighted random selection problem; the candidate solved two fully and one partially and was shortlisted.
Covered · Dynamic programming, Arrays & strings
2
DSATechnical
A binary-search framing that the interviewer redirected toward a sliding window after a long discussion, then a grid traversal problem answered as pseudocode when time ran out.
Covered · Arrays & strings, Graphs, Data structures & algorithms
3
DSA and CS fundamentalsTechnical
Threads and object-oriented concepts alongside a medium tree-width problem.
Covered · Concurrency, OOP, Trees
4
HRBehavioral
Heavy cross-examination of the current project to test hands-on depth, plus behavioural questions.
A two-hour assessment applying data-science techniques to scenario problems, mixing coding with data manipulation and statistical reasoning.
Covered · Data structures & algorithms, SQL, Aptitude
2
Statistics and machine learningTechnical
Forty-five minutes on statistical models, hypothesis testing, regression and visualisation, then tree-based and boosting algorithms, closing with a scenario where the candidate proposed a modelling approach.
Two DSA problems, one binary-search-on-answer and one tree modification task, plus multiple-choice questions on data manipulation and statistics.
Covered · Trees, Data structures & algorithms, Aptitude
2
Statistics and machine learningTechnical
Forty-five minutes on hypothesis testing, regression and visualisation, then ensembles, gradient descent and the bias-variance trade-off, ending in a short case where a modelling approach had to be chosen.
Covered · Statistics, Machine learning, Case study
3
Experimental design and A/B testingTechnical
Metric choice and experiment design, with the discussion pushed toward the practical difficulties of running tests at large scale.
Covered · A/B testing, Statistics
4
Hiring managerHiring manager
A thirty-minute values, mission and career-goals conversation.
An introduction and project story, then live SQL from subqueries and joins up to self-joins and unpivoting, Python and pandas output-prediction questions, SQL theory contrasts, and a business prompt on judging an Uber Eats city launch. A separate one-hour timed assessment covered quant, verbal, SQL and hypothesis testing in a provided spreadsheet.
Covered · SQL, Python, Statistics, Case study
2
Technical and businessCase study
A guesstimate of Uber's ride share in the candidate's own city, then a case on which KPIs would decide within three months whether to keep operating in a new city.
Covered · Market sizing, Case study, Domain knowledge
3
Panel on business understandingPanel
Two interviewers probing the candidate's measurable impact and how their analysis changed decisions, then a case on classifying a driver as good or bad, choosing four metrics and applying them to given values.
Covered · Case study, Domain knowledge, Past projects
4
Panel on stakeholder managementPanel
An informal get-to-know-you opening, then stories about conflicting stakeholders and unmeetable deadlines, followed by a case on driver retention in a new city and a critique of a rides-per-hour metric.
Covered · Conflict handling, STAR stories, Case study
A week to submit a product proposal as slides. Uber's instructions spelled out the evaluation criteria and required components, including MVP definition, prioritisation and a go-to-market plan; the author submitted annotated paper wireframes.
Covered · Product sense, Case study
2
Product interviewCase study
Forty-five minutes split into a short behavioural opening, a thirty-five-minute product case and time for questions.
Covered · Product sense, Behavioral
3
Product interviewCase study
A second interview in the same format with a different product manager.
Covered · Product sense, Behavioral
4
Virtual onsite (four interviews)Panel
Four forty-five-minute sessions in one day with two product managers, an engineer and a UX designer. The engineering session was a collaborative system-design discussion where the PM candidate was expected to co-design rather than solve alone; the designer session was behavioural, focused on how the candidate has worked with design partners.
Covered · Product sense, System design, Behavioral
Submitted through HackerRank about two weeks after applying, walking through a full product lifecycle with roughly six hours of recommended effort. The author notes friends were cut for missing one of the stated criteria.
Covered · Product sense, Case study
2
Product phone interviewCase study
A forty-five-minute product case run by a PM or current APM; the author stresses building personas for both sides of the marketplace, riders and drivers.
Covered · Product sense, Domain knowledge
3
Analytical phone interviewCase study
Working through Uber-specific metric scenarios and being explicit about how any change moves driver supply, rider demand and the company's economics.
Three hours at the San Francisco office: two PM sessions repeating the product and analytical formats, a product-designer session about how the candidate has handled user journeys in past work, and an engineer session testing whether the candidate can explain a genuinely technical project they owned.
What is the interview process like at Uber Technologies?
Uber runs a highly standardised, named-round loop that candidates describe almost identically across geographies: a timed online assessment (CodeSignal, HackerRank or Codility), a screening or BPS coding call that acts as the elimination gate, then a same-day set of one-hour rounds with fixed labels - Algorithms & Data Structures, Depth in Specialization (production-quality machine coding), Design & Architecture on a new problem, Design & Architecture on a previously solved problem, and a Collaboration/Leadership or bar-raiser conversation with a manager. Recruiters brief candidates on the round names and send prep material in advance, so the loop is unusually predictable; the bar inside it is not, with several writers reporting rejections for readability, missed edge cases or a project story judged not to be at Uber scale. Non-engineering loops diverge sharply: product management runs a multi-hour take-home plus a group-style product case day with PM, engineer and UX-designer interviewers, and analytics roles pair SQL/Python screening with city-launch and marketplace case studies.
What topics does Uber Technologies test in interviews?
Commonly reported topics include Data structures & algorithms, Dynamic programming, Graphs, Trees, Arrays & strings, System design.
These guides summarize public, first-hand interview experiences shared by candidates. They describe the shape of each loop and the topics that came up, not a leaked question bank, and every experience links back to its original source. Processes change often; treat this as directional prep, not a script.
