CloudSEK | SDE-1 | Rejected | Bad Experience
Summary
I had a rather chaotic and unexpected SDE-1 interview experience with CloudSEK, where I was primarily asked irrelevant and advanced questions by a Security Engineer, despite applying for a fresher role. This misalignment, coupled with poor interviewer conduct, ultimately led to my rejection.
Full Experience
Introduction
I recently interviewed with CloudSEK for a Software Engineer role targeting freshers (0+ years of experience), as I'm a 2023 graduate currently working as an SDE - Infrastructure. After a phone screening and a successful first technical round with an SDE-2, things took a turn.
Technical Round 2 - The Chaos Begins
This round was supposed to focus on my work experience, DSA, and core CS concepts, but to my surprise, the interviewer was a Security Engineer. He joined the meeting 15 minutes late and didn't even ask for an introduction, which set an odd tone.
These are some of the questions he asked, in addition to other irrelevant ones:
Q: "What do you know about APIs, any practical experience?"
I explained that while I hadn't directly implemented APIs, I understood the concepts well. Without acknowledging my response, he immediately asked me to open a code editor and posed this problem:
Q: "You are given a list of URLs. You need to make requests to these URLs, retrieve the metadata, and store it in a JSON file. Also, how can you make the process faster?"
I was baffled, as this felt more like a use-case-specific problem than a typical DSA question. I politely offered to discuss the approach and write pseudocode. His response was a humiliating, "But you said you know the basics of APIs." He quickly diverted to multithreading, concurrency, and asynchronous I/O, topics I hadn't worked on, and wanted me to code instances of async I/O.
Q: "Let’s say you have 2 million records, and the read is slow. What would you do to improve it?"
I suggested implementing indexing on the relevant column. Instead of acknowledging, he pushed further:
Q: "Let’s say you have two indexed columns, A and B, and a compound index on A+B. From where would the fetching be faster?"
I had never heard the term 'compound indexing' before, but I answered his questions about indexing confidently.
Q: "Where exactly is the index created? Could you show diagrammatically where an index resides in a DB?"
Fortunately, I was able to answer these well. But then came another unexpected turn:
Q: "What do you know about GO? Talk to me about the low level working of GO"?
I honestly stated I hadn't worked with GO, yet he proceeded to give me a GO-related problem. I did my best to discuss it by relating concepts to Python and C++.
Next, an irrelevant question:
Q: "You have different sharded machines (machine A, B, C, etc.) with sharded DBs, and you're unable to retrieve data from machine B. What would you do to ensure data is retrievable from machine B?"
I explained my troubleshooting approach, but again, his reaction was neutral. The interview ended with a random question about Kubernetes and no chance for me to ask questions. He simply said, "HR will reach out to you."
Conclusion
The interview lasted 90 minutes. I was constantly bombarded with questions irrelevant to the role and my experience, often taunted with "What have you worked on then?" My resume was largely ignored. Apart from brief discussions on Kubernetes, Docker, and Networking, the conversation felt completely off-track, more suited for a senior role. I later learned from HR that I didn't pass, which she found surprising. Despite her assurance of updates, I never heard back, suggesting they might have realized the interview was flawed but didn't want to admit it.
Interview Questions (7)
What do you know about APIs, any practical experience?
You are given a list of URLs. You need to make requests to these URLs, retrieve the metadata, and store it in a JSON file. Also, how can you make the process faster?
Let’s say you have 2 million records, and the read is slow. What would you do to improve it?
Let’s say you have two indexed columns, A and B, and a compound index on A+B. From where would the fetching be faster?
Where exactly is the index created? Could you show diagrammatically where an index resides in a DB?
What do you know about GO? Talk to me about the low level working of GO?
You have different sharded machines (machine A, B, C, etc.) with sharded DBs, and you're unable to retrieve data from machine B. What would you do to ensure data is retrievable from machine B?