SDE-2 Interview Experience at CRED (Rejected) | Dec 2025
Summary
I interviewed for an SDE-2 role at CRED in December 2025 and was ultimately rejected. The interview process included initial screening, a comprehensive system design and low-level design round with a take-home assignment, and a hiring manager round, where my shallow understanding of fundamental concepts like LSM Trees, SSTables, and AI attention mechanisms led to my rejection.
Full Experience
SDE-2 Interview Experience at CRED (Rejected) | Dec 2025
Recently, I interviewed for an SDE-2 role at CRED. Sharing my experience in detail—both to help others and to reflect on what I learned.
How It Started
I received a call from HR mentioning that they got my profile via Instahyre. After a brief discussion about my background, they scheduled the interview process.
Round 1: Initial Screening (Online)
Duration: ~30 minutes (ended in ~20 minutes)
This was more like a pseudo hiring-manager screening round.
Discussion topics: • My current tech stack and how I use it in production • How I use AI tools in day-to-day engineering work • High-level discussion about my work experience
The round felt conversational and was primarily to assess fit and breadth, not depth.
Outcome: Next day, HR confirmed that I was moving to the next round.
⸻
Round 2: Design + LLD (Face-to-Face, Onsite at CRED Office)
Duration: ~2 hours Focus: Low-Level Design + Problem Solving
This round was aligned with CRED’s AI-Native Design & Coding expectations .
Problem Statement
Design a Flight Booking System (similar to MakeMyTrip).
Expectations & Discussion Areas
I was expected to lead the entire design, starting from scratch:
-
Requirements • Functional & non-functional requirements • Clear scoping and assumptions
-
Low Level Design • Core Domain & Entities • Database schema design and relationships
-
Flight Search • One-way and round-trip • Connecting flights with at most 1 stop • Search performance considerations
-
Booking Flow • Seat locking and seat booking • Price fluctuation handling (each seat has its own price) • Booking segments and consistency
-
Multi-Vendor Problem (How to handle ?) • You are one booking platform • Other platforms (e.g., MakeMyTrip, Goibibo) also sell the same seats (My Answer : This will handle at airline end. Airline will lock on platforms behalf. (Interviewer convinced.)) • How to prevent double booking across platforms
-
High-Level Architecture • Services • Database • Caching (Redis) • Messaging (Kafka)
⸻
Take-Home Assignment
After the round, the interviewer asked me to: • Submit a clear design document • Build a working prototype within 24 hours • Ensure all mentioned components are connected (DB, Redis, Kafka, etc.)
AI usage was explicitly allowed.
I implemented the system using Docker and Docker Compose, wired all components together, and submitted the project within the deadline.
⸻
**Round 3: Hiring Manager Round (Face-to-Face, Onsite CRED Office) ** Duration: ~1 hours (Went till ~1.5 hours) Panel: 2 interviewers
Part 1 (First ~1 hour): Project & Experience Deep Dive • Detailed discussion on my submitted project • Architecture choices • Trade-offs • Past work experience
This part went well and felt like a healthy technical discussion.
⸻
Part 2 (Last ~30 minutes): System & Fundamentals Deep Dive
This is where things got difficult.
Questions That Exposed Gaps 1. I mentioned database indexing → deep dive into LSM Trees • I couldn’t explain it clearly 2. I mentioned BigTable → asked about SSTables • Again, no clear answer 3. AI fundamentals: • “Have you read Attention Is All You Need?” • What is attention and why it matters? 4. Learning approach: • Do you read research papers from OpenAI, Anthropic, Google, etc.? • How do you understand how AI systems work internally?
These questions were chained, each going deeper based on my previous answers and worked i have done.
⸻
Feedback & Verdict
The interviewer asked me to rate my own interview out of 10. I said 6/10.
At the end, the interviewer was transparent and shared negative feedback, mainly around: • Shallow understanding of fundamentals • Mentioning concepts without deep clarity
Final Result: ❌ Rejected
⸻
Key Takeaway (Most Important)
Surface-level knowledge will kill you in senior interviews.
If you mention: • LSM Trees → you must know how they work • BigTable → you must understand SSTables • Attention → you must know why it exists and how it works
Name-dropping concepts without depth backfires hard.
⸻
What I Learned • Senior interviews are not about what you’ve used, but how deeply you understand it • If you’re learning something: • Go end-to-end • Read internals • Understand trade-offs
This interview was a humbling but extremely valuable experience—and a strong reminder to go deep, not wide.
Interview Questions (6)
Design a Flight Booking System (similar to MakeMyTrip).
Expectations & Discussion Areas
I was expected to lead the entire design, starting from scratch:
-
Requirements • Functional & non-functional requirements • Clear scoping and assumptions
-
Low Level Design • Core Domain & Entities • Database schema design and relationships
-
Flight Search • One-way and round-trip • Connecting flights with at most 1 stop • Search performance considerations
-
Booking Flow • Seat locking and seat booking • Price fluctuation handling (each seat has its own price) • Booking segments and consistency
-
Multi-Vendor Problem (How to handle ?) • You are one booking platform • Other platforms (e.g., MakeMyTrip, Goibibo) also sell the same seats • How to prevent double booking across platforms
-
High-Level Architecture • Services • Database • Caching (Redis) • Messaging (Kafka)
After the round, the interviewer asked me to: • Submit a clear design document • Build a working prototype within 24 hours • Ensure all mentioned components are connected (DB, Redis, Kafka, etc.)
AI usage was explicitly allowed.
Deep dive into LSM Trees, triggered by mentioning database indexing.
Deep dive into SSTables, in the context of BigTable.
Questions about AI fundamentals: "Have you read Attention Is All You Need?" What is attention and why it matters?
Discussion on learning approach: Do you read research papers from OpenAI, Anthropic, Google, etc.? How do you understand how AI systems work internally?