Razorpay Interview Questions 2026: SDE and Data Roles, Coding, and Fintech Domain Knowledge

Razorpay SDE interview process 2026:

1. Resume Screen: Razorpay looks for: product company experience (Flipkart, Swiggy, Amazon India, or well-funded startups), quantified impact in experience bullets (not 'developed APIs' but 'built a checkout API handling 50K transactions/day'), relevant side projects, IIT/NIT background (helps but is not required), and a strong GitHub profile.

2. Online Assessment (30-90 minutes): - 2 coding problems, LeetCode medium-hard difficulty - Common topics: interval problems, priority queues, graphs, sliding window - For experienced hires: sometimes replaced by a take-home design exercise

3. Technical Coding Rounds (2 rounds, 45-60 minutes each): Round 1: arrays, strings, trees (LeetCode medium) Round 2: graphs, DP, or concurrency (LeetCode medium-hard for SDE-2+)

4. System Design Round (SDE-2 and above, 45-60 minutes): Expected depth: distributed systems, data modelling, API design, scalability trade-offs. Fintech-specific: idempotency, ACID transactions, payment state machines, fraud signals.

5. Technical Culture Fit (30-45 minutes): Razorpay explicitly evaluates: Curiosity (do you ask why, not just how?), Ownership (do you fix problems or report them?), High standards (do you care about code quality and system reliability?), Customer obsession (do you think about the downstream impact of your decisions on merchants and their end users?).

1. Hiring Manager Round: what specifically you would contribute to the team.

Razorpay most common coding topics:

Financial computation problems (Razorpay-specific style):

• 'Given a list of transactions, find the minimum number of transactions to settle all debts' (graph problem: compute net balance per person, then greedily match largest creditor with largest debtor)
• 'Design a rate limiter function that limits N requests per second per client' (sliding window counter using a deque or token bucket)
• 'Given a list of payment events, find the maximum payment amount within any 5-minute window' (sliding window maximum)

Standard coding topics:

• Arrays: merge intervals, trapping rain water, jump game II
• Trees: binary tree right side view, serialize/deserialize
• Graphs: course schedule II (topological sort), word ladder (BFS)
• DP: coin change (minimum coins), longest increasing subsequence
• Priority queue: K closest points to origin, find the running median (two heaps), meeting rooms II (minimum rooms needed)

Concurrency (SDE-2+):

• 'Implement a thread-safe LRU cache in Java'
• 'What is the difference between AtomicInteger and synchronized in Java and when would you use each?'
• 'How do you implement a producer-consumer pattern?'

Razorpay system design questions:

1. 'Design a payment gateway': Key concerns: idempotency (each payment request has a unique idempotency key; duplicate requests return the existing response, not a second charge), payment state machine (initiated, processing, authorised, captured, failed, refunded; each transition is immutable in the database), low latency (P99 under 500ms for the customer-facing checkout flow), reliability (circuit breaker for network provider calls; fallback to a secondary provider on failure).

2. 'Design a fraud detection system at Razorpay scale': Feature engineering: amount deviation from merchant's historical average, device fingerprint novelty, time-since-last-transaction, geographic velocity (same card used in two distant locations within 10 minutes). Model serving: gradient boosting model scored per transaction in under 100ms P99 via a cached feature store (Redis). Decision: low-risk proceed, medium-risk request OTP, high-risk decline. Feedback loop: confirmed fraud labels from chargeback data retrain the model weekly.

3. 'Design a ledger system for merchant payouts': Double-entry bookkeeping: every transaction has two entries (debit from the pool account, credit to the merchant account); the sum of all debits must equal the sum of all credits at all times. Immutability: ledger entries are never updated or deleted; corrections are made via reversal entries. Settlement: batch job at the end of the day (T+1 or T+2) aggregates net merchant balances and initiates NEFT/IMPS/UPI transfers.

What Razorpay looks for beyond technical skills:

1. Payment domain depth: Razorpay engineers are expected to understand: UPI (Unified Payments Interface: NPCI's real-time payments system; built on IMPS; linked to VPA (Virtual Payment Address) not bank account numbers directly), payment rails (Visa/Mastercard for card payments; RBI-regulated card networks; interchange fee model), NACH (National Automated Clearing House: for recurring debit mandates, EMI collection), RBI compliance (PCI-DSS, data localisation requirements for payment data in India). This is not tested explicitly in the interview but signals domain commitment.

2. Production ownership: Razorpay's culture: the engineer who writes the code is also responsible for it in production. 'You build it, you run it.' Be prepared to discuss: how you have been paged or alerted for production issues, what on-call means to you, how you have improved observability or runbooks for your team.

3. Code quality standards: Razorpay is known within the Indian engineering community for high code review standards. Engineers are expected to write comprehensive tests, document non-obvious decisions, and leave the codebase cleaner than they found it.

4. Collaboration across functions: Razorpay's engineering teams work closely with product, operations, and merchant support. Be ready to discuss: how you communicate trade-offs to non-technical stakeholders, how you have collaborated with PM to define APIs or product behaviour.

Razorpay total compensation (2024-25, Bangalore):

SDE-1 (0-3 years experience): base 20-30 LPA + RSUs 8-15 LPA annualised = 28-45 LPA TC SDE-2 (3-6 years): base 30-50 LPA + RSUs 15-30 LPA = 45-80 LPA TC Senior SDE (6-10 years): base 50-80 LPA + RSUs 30-60 LPA = 80-140 LPA TC Staff SDE (10+ years): 80+ LPA base + 50-80 LPA RSUs = 130-200 LPA TC

Note: Razorpay is a pre-IPO company (Series F funded as of 2025). RSUs are company stock; liquidity depends on an IPO or secondary sale event. The valuation was approximately USD 7.5 billion in 2021; market conditions affect actual RSU value at liquidation.

Why engineers choose Razorpay:

• 5-10x salary vs IT services companies
• Genuinely hard engineering problems at national payment scale
• Engineers own production systems (no separate SRE/ops team for most teams)
• Brand recognition in the Indian startup ecosystem (opens doors to other fintech companies, FAANG India, or founding your own startup)

Career from Razorpay: After 3-5 years at Razorpay, engineers typically move to: FAANG India (Amazon, Google, Microsoft pay more but Razorpay is a respected feeder), PhonePe, CRED, or Juspay for lateral moves, or global fintech companies (Stripe, Adyen) for international opportunities.