CRED and PhonePe are two of India's most prominent fintech companies and both are known for high engineering culture. CRED serves India's premium credit card users; PhonePe is India's largest UPI payment platform. Both offer strong compensation and challenging engineering problems. This guide covers the complete interview process and question bank for both companies in 2026.
CRED SDE interview process 2026:
1. Resume screen: CRED is selective. The minimum bar for experienced hires is 2-3 years of strong product company experience. For freshers: IIT/NIT with exceptional competitive programming credentials or a strong open source portfolio.
2. Online Assessment or Take-Home: - Experienced hires (2+ years): often a take-home coding assignment (2-4 hours): design and implement a small system or feature with clean, testable code - Freshers (rare direct hiring): 2-3 LeetCode medium-hard problems in 90 minutes
3. Technical Coding Rounds (2-3 rounds, 60 minutes each): CRED asks harder algorithmic problems than most Indian unicorns: - Graph algorithms: Dijkstra's for weighted shortest path, minimum spanning tree, network flow - Advanced DP: bitmask DP, digit DP, interval DP - Concurrency: design a thread-safe data structure; race condition identification - System design basics in coding rounds: 'implement an LRU cache', 'implement a rate limiter'
4. System Design Round (SDE-2 and above): - 'Design CRED's bill payment reminder and notification system' - 'Design a credit score monitoring system' - 'Design CRED's rewards point system with expiry and redemption logic'
5. Engineering Culture Fit: CRED is known for extremely high code quality standards; interviewers explicitly evaluate: do you write tests? do you think about edge cases before being asked? do you care about naming and readability?
PhonePe SDE interview process 2026:
1. Resume screen: PhonePe values: product company experience, projects with quantified scale, and payment or fintech domain knowledge.
2. Online Assessment (60-90 minutes): - 2 coding problems (LeetCode medium-hard) - Sometimes includes a system design problem description to analyse in writing (for SDE-2+)
3. Technical Coding Rounds (2-3 rounds, 45-60 minutes each): - Arrays and strings: median of a stream (two heaps), interval problems - Trees: build and query a segment tree or Fenwick tree (SDE-2+) - Concurrency: 'Implement a thread-safe bounded queue', 'Explain Java's volatile vs synchronized' - Payment-domain coding: 'given a list of transactions, compute each user's balance; detect if any account is overdrawn'
4. System Design Round (SDE-2 and above): - 'Design UPI payment system at 500M users' - 'Design PhonePe's transaction history service' - 'Design a real-time balance update system with consistency guarantees' - 'Design a merchant dashboard with near-real-time analytics'
5. Behavioural Round: PhonePe values ownership and customer obsession (influenced by its time as a Flipkart subsidiary, which was influenced by Amazon). Prepare STAR stories for: taking ownership of a production issue, driving a cross-functional initiative, disagreeing with a technical decision and how you handled it.
CRED system design questions:
1. 'Design CRED's bill payment reminder system': User context: CRED users link their credit cards; the system knows their billing cycle and due dates. Reminder pipeline: scheduled job checks all users 7 days, 3 days, and 1 day before due date; generates a personalised notification payload (bill amount, minimum due, due date). Delivery: push notification (FCM for Android, APNs for iOS), SMS for users who opt in, in-app notification (WebSocket for open app; fetched on next open for background users). Deduplication: track sent notifications per user per billing cycle in Redis with TTL to avoid re-sending on retry.
2. 'Design a credit score monitoring system for CRED': Data source: credit bureau (CIBIL, Experian, Equifax) provides credit reports via API; CRED has a license to pull reports. Scheduling: monthly pull per user on a rotating schedule (not all on the 1st to avoid thundering herd). Change detection: compare new score to previous; if delta exceeds threshold (score dropped 20+ points), trigger an alert. Score explanation: LLM-generated personalised explanation of what factors changed (new inquiry, utilisation increase). Storage: score history per user in a time-series table with monthly granularity.
PhonePe system design'Design UPI payment flow': Initiation: payer's app sends payment request to PhonePe's PSP (Payment Service Provider) server. Resolution: PhonePe resolves the payee's VPA to their bank account via NPCI's UPIID lookup. Authorisation: PhonePe sends a collect request to the payer's bank via NPCI; the payer's bank sends a PIN challenge; payer enters PIN on their bank app. Settlement: NPCI routes the fund transfer; payer's bank debits, payee's bank credits. Status: webhook to both payer and payee apps on completion.
CRED engineering culture- Code quality is a first-class concern: every PR has a thorough review; engineers are expected to write tests before asking for review - Architecture decisions are documented: RFCs (Request for Comments) are common before major system changes - 'Do it right': CRED is known for moving more deliberately than most Indian startups; they do not ship something they are not proud of - Tech stack: Kotlin for backend (JVM-based, strong type safety), Golang for performance-critical services, React Native for mobile - Engineering blog: CRED has a well-regarded engineering blog; reading it before the interview signals genuine interest
PhonePe engineering culture: - Scale obsession: PhonePe handles 500M+ users and trillions of INR in transaction volume; engineers are expected to think about scale from day 1 - Structured process: Walmart influence means more process (design reviews, incident post-mortems, SLA tracking) than most Indian startups - On-call culture: PhonePe has a formal on-call rotation; engineers own their services in production - Tech stack: Java and Spring Boot for core payment services, Go for high-throughput services, React Native and Flutter for mobile, Apache Kafka for event streaming, Apache Flink for real-time stream processing
CRED total compensation (2024-25, Bangalore): - SDE-1: base 30-40 LPA + RSUs 10-20 LPA annualised = 40-60 LPA TC - SDE-2: base 45-65 LPA + RSUs 20-35 LPA = 65-100 LPA TC - Senior SDE: base 65-90 LPA + RSUs 35-60 LPA = 100-150 LPA TC
CRED is pre-IPO (Series F funded, valued at approximately USD 6.4 billion in 2021). RSUs are company stock; liquidity depends on IPO or secondary sale.
PhonePe total compensation (2024-25, Bangalore and Pune): - SDE-1: base 25-35 LPA + RSUs 10-20 LPA = 35-55 LPA TC - SDE-2: base 35-55 LPA + RSUs 20-35 LPA = 55-90 LPA TC - Senior SDE: base 55-80 LPA + RSUs 35-60 LPA = 90-140 LPA TC
PhonePe is Walmart-backed (Walmart owns approximately 85%); not listed publicly as of 2025. RSU liquidity similarly depends on a future listing or buyback programme.
Both offer Joining Bonustypically 20-50 LPA as a one-time payment to compensate for unvested stock being forfeited at the previous company.