My First Technical Interview (AlgoUniversity)

Today I gave my first ever technical interview for an SDE intern role at AlgoUniversity.

There were 2 DSA questions and 2 CS fundamentals (OS + OOP).

This is a straightforward breakdown of what was asked, what I thought, what was right, and what I missed.

DSA Question 1 — Search in Infinite Sorted ArrayCopied!

Problem

Sorted array (monotonically increasing, unique).
Length unknown (conceptually infinite).
Given a target → return index or -1.

Example:

arr = [2, 5, 9, 14, 21, 37, 58, 73, 91, ...]
target = 37 → 5

What I thoughtCopied!

Binary search is ideal since sorted.
But right boundary unknown.

So I tried expanding range exponentially:

while (arr[last] < target)
    last *= 2;

I actually wrote this part correctly.

But I embedded it inside a binary search loop instead of treating it as a separate phase:

while (first < last) {
    while (arr[last] < target)
        last *= 2;

    mid = (first + last) / 2;
    // binary search logic
}

What was rightCopied!

• Realized boundary must be discovered
• Used exponential growth (correct idea)

What was wrongCopied!

• Mixed range expansion with binary search
• Didn’t separate phases

Correct structureCopied!

1. Expand range exponentially
2. Binary search inside range

LearningCopied!

When array size unknown → exponential search first, binary search later.
Structure matters as much as idea.

DSA Question 2 — Minimum Cost Grid PathCopied!

Problem

n×n grid with costs.
Start: (0,0)
End: (n−1,n−1)
Move in 4 directions.
Minimize total cost.

Example:

1 3 1
2 1 4
3 2 1

What I thoughtCopied!

Grid → graph
Cell → node
Neighbors → edges
Cost → weight

So I chose Dijkstra.

Started with direction arrays:

dirX = {1,0,0,-1}
dirY = {0,1,-1,0}

Then I drifted toward BFS + “pick smallest each stage”.

What was rightCopied!

• Correct modeling
• Correct algorithm choice (Dijkstra)

What was wrongCopied!

• Forgot priority queue implementation
• Couldn’t recall Dijkstra mechanics

LearningCopied!

Knowing algorithm name ≠ implementing it.
Need recall fluency for core algorithms (Dijkstra, BFS, DFS).

OS Question — 2GB RAM Running 5GB GameCopied!

Question

How can a 2GB RAM laptop run a 5GB game?

My answerCopied!

Swap memory on disk acting like extra RAM.

EvaluationCopied!

Correct concept.

Full idea: virtual memory → inactive pages moved to disk.

LearningCopied!

OS fundamentals are mostly clear.

OOP Question — Virtual DestructorsCopied!

Question

What is a virtual destructor and why important?

My answerCopied!

I misunderstood and talked about auto type deduction.
So this was wrong.

Correct conceptCopied!

If deleting via base pointer:

Base* b = new Derived();
delete b;

Without virtual destructor → only base destroyed.
Derived cleanup skipped → leak.

LearningCopied!

Need stronger OOP fundamentals recall.

Interview FeedbackCopied!

Interviewer said:

• Improve DSA
• Talk more and discuss approach
• He would give hints if I interacted more

Overall LearningCopied!

• I was often structurally close in DSA
• Main gaps: recall + articulation
• Need practice implementing core graph algorithms
• Need to think aloud in interviews

First technical interview → clear direction now.