Carwale SDE-1 Interview Experience

Round 1: DSA + Problem Solving (Technical Coding Round)

• Duration: 60 minutes
• Mode: Remote (Live Coding)
• Difficulty: Medium

Questions Asked:

1. Search in Rotated Sorted Array
2. Print Top View of Binary Tree
3. Array-based problem (implementation-focused)

My Experience:

The round began with a quick self-introduction, followed by a classic DSA question — "Search in a Rotated Sorted Array". I was initially stuck on some edge cases and needed a few hints to get back on track. Once reminded about the possible edge conditions, I corrected my approach and completed the solution and Write a working code.

The interviewer then asked a conceptual tree-based problem: Print Top View of Binary Tree, where I was able to come up with a working pseudocode after a few nudges.

The final question involved an array manipulation task, where I successfully wrote a complete working solution without much support.

Throughout the round, the focus was on logical thinking, debugging on the fly and edge case handling.

Verdict:

DSA-focused candidate with strong problem-solving skills. Needs to improve handling of edge cases independently and develop more confidence under pressure.

Round 2: Project Deep Dive + System Thinking

• Duration: 60 minutes
• Mode: Remote
• Difficulty: Medium

Discussion Topics:

• Deep dive into Hackathon Project (AI + avatar lip-syncing)
• Limitations, bottlenecks (RAM usage)
• Scaling strategies and cost optimization
• Caching strategies and implementation

My Experience:

We started by discussing my hackathon project, which involved facial analysis, AI-driven question answering and lip-syncing with an avatar. The interviewer asked in-depth questions about the tech stack and implementation, especially around performance and scalability.

I identified RAM as a bottleneck and proposed horizontal scaling. When asked about cost optimization (e.g., handling idle servers at night), I mentioned shutting down machines but struggled to detail how to automate this — a clear learning point.

The discussion moved toward caching strategies. I correctly identified redundant computations and suggested a shared cache across servers. However, I initially confused CPU L1 cache with application-level caching.

Eventually, I proposed using hashmaps and browser-based localStorage for client-side caching, especially when the same user accesses repeated data.

Despite some confusion and slow starts, I was able to think through the problems and reach practical solutions.

Verdict:

Shows strong project ownership and real-world reasoning. Needs to deepen understanding of system design concepts, cost optimization techniques, and caching architectures.

Round 3: System Design + Behavioral Insight

• Duration: 60 minutes
• Mode: Remote
• Difficulty: Medium

Topics Covered:

• Scaling and optimization
• Handling concurrency
• Edge case considerations
• Behavioral Q&A (growth mindset, adaptability)

My Experience:

This round felt like a blend of system design discussion and behavioral analysis.

We discussed:

• What happens if multiple users ask the same question?
  → I recognized the need for deduplication and caching.

• How to cache at different layers?
  → I suggested using server-side hashmap cache and browser localStorage for repeated user access.

• How to implement such a cache system?
  → I proposed a simple hashmap-based cache with key-value storage.

Behavioral Questions:

• What if your solution doesn’t scale?
• What would you do if your team disagrees on architecture?

While I didn’t have all the answers right away, I remained calm, clarified questions and reasoned my way through.

Verdict:

Logical and thoughtful problem-solver. Demonstrates growth potential and willingness to learn. Would benefit from mentoring in scalable system design.