Company: Oracle (CGIU) - 6M Intern + FTE
Role: Associate Software Developer
Date: August '25
Mode: On-campus
Verdict: Selected

Round 1 – Online Assessment (OA)

The OA had four sections covering aptitude, reading comprehension, probability, and core CS fundamentals like DBMS, OOP, SQL, OS, and Networking. There were also two coding problems:

1. Minimize Maximum of Array
Given an array of positive integers. In each operation, you could pick an index i (2 ≤ i ≤ N) and a value x (1 ≤ x ≤ arr[i]), then do:

• arr[i-1] = arr[i-1] + x

• arr[i] = arr[i] - x

The goal was to minimize the maximum value in the array. I solved it using binary search on the possible maximum combined with a greedy check function that pushed surplus values left.

2. JSON Diff Tool
Develop a simple service to compare two JSON objects (up to size 1e5) and identify differences in their key-value pairs.

Round 2 – Technical Interview 1

This round started with introductions and some resume-based questions. The interviewer asked me about my internship experience and technical concepts I had worked on.

Resume-based Questions:

• What are messaging queues? Give examples.

• What is AOF and RDB in Redis? Explain the differences.

• Explain the architecture and details of the projects on my resume.

DSA Question: Longest Increasing Subsequence (LIS)

• I first explained the brute force recursive approach.

• Then I optimized it using DP tabulation (O(n²)).

• Finally, I discussed further optimization using binary search with extra space (O(n log n)) and wrote pseudocode in C++.

Round 3 – Technical Interview 2

The round started with a conceptual question:
Why do we need Redis when programming languages already have hashmaps?

• I explained that Redis provides persistence, rich data structures, and additional features that go beyond simple in-memory maps.

DSA Questions:

1. Find the nth node from the back of a linked list – solved using two pointers.

2. Find the Lowest Common Ancestor (LCA) in a binary tree – solved using DFS.

3. Find all pairs of numbers whose sum is divisible by 60 – solved using complements and hashing.

Round 4 – Technical Interview 3

This was a short round with one question:

Question: Count all substrings in a very long string of digits that represent powers of 2. The string was too large to fit in RAM and resided on disk.

My Approach:
I processed the string in chunks, making sure to handle overlaps at chunk boundaries. I precomputed all powers of 2 as strings and stored them in a hash set for fast lookup. Then, I slid a window over each chunk and checked substrings against the set. From a system design perspective, I treated it like a small pipeline: a disk reader loads chunks, a processor extracts and checks substrings, and an aggregator combines counts. Multiple chunks could be processed in parallel to make it scalable...

Round 5 – Hiring Manager (HM) Round

This round included a mix of puzzle and behavioral questions.

Puzzle:

• You have a 3L jug and a 5L jug with infinite water supply. Measure exactly 4 liters without any measuring marks.

Behavioral / Scenario Questions:

• Where do you see yourself in 5 years?

• If you were a tech lead, how would you manage your team?

• How would you handle a situation where a team member is absent, manpower is short, and a deadline is approaching?

• Do you have plans for pursuing MS in the future?

• What are some real-life applications of data structures?