Expect a streaming round on computing surge-relevant supply/demand metrics per geohash from trip events with bounded latency and out-of-order handling, a SQL round on percentile metrics (median trip duration per city-hour) and scaling to billions of rows with approximate algorithms, and a platform round on designing a self-serve semantic/metrics layer for consistent metric definitions. Interviewers reward precise reasoning about windowing, watermarks, geospatial bucketing, and metric governance rather than generic ETL. Uber pioneered internal metrics platforms, so governance thinking stands out.
Recruitment process
- Recruiter screen and technical pre-screen
- SQL and data-manipulation round
- Streaming pipeline and platform design round
- Hiring-manager and behavioural round, then offer
Interview rounds
- Round 1 (45-60 min): SQL round on percentiles and large-scale analytical queries
- Round 2 (60 min): streaming pipeline design round with windowing and out-of-order handling
- Round 3 (45-60 min): metrics-platform and semantic-layer design round
- Round 4 (45 min): behavioural and hiring-manager round
Stream and standardise framework
- Design low-latency streaming with watermarks and geospatial bucketing, compute percentiles at scale with approximate algorithms, and standardise metrics via a semantic layer
Practice questions
- 01Practise
Reverse a linked list in O(1) extra space. Then explain when you'd use this in production.
Standard - 02Practise
What is the difference between a class and an object? Give an example.
Warm-upTCS Ninja/Digital tracks lean heavily on CS fundamentals: OOP, OS, DBMS, SQL basics. Expect rapid-fire Q's, not deep dives. - 03Practise
Write a SQL query to find the second-highest salary in a department.
Warm-up - 04Practise
What is normalisation in databases? When would you de-normalise?
Warm-up - 05Practise
What's the difference between a stack and a queue? Give a real-world use case for each.
Warm-up - 06Practise
Given a stream of integers, design a data structure that returns the median in O(log n).
Intense - 07Practise
You have a service that's slow only on Tuesdays. Walk me through how you'd debug.
StandardGoogleyness: structured debugging, hypothesis-driven, instrumentation-first. - 08Practise
Here's 200 lines of payment-processing code with several intentional bugs. Find them and explain the production impact of each.
IntenseStripe's signature 'bug bash' round. Off-by-one, race conditions, wrong error handling, edge cases in money math. Communicating impact > finding all bugs. - 09Practise
Explain webhooks vs. polling for a developer who has never built a payment integration. You have 4 sentences.
StandardStripe values writing clarity above almost everything. Length-constrained explanation tests prioritisation of ideas. - 10Practise
Walk me through Razorpay's fraud-detection model. How would you reduce false positives without increasing chargeback losses?
IntenseRazorpay ML rounds tie model decisions to ₹ impact, chargeback cost vs. blocked-good-txn cost. Pure ML answers without business framing fail.
