Cognam Technologies Interview Experience for FTE

First Round: MCQ and Coding Round with different sections and each section had a different time limit

1. Section 1: It consists of 10 aptitude MCQ questions to be solved in 10 mins.
2. Section 2: It consists of 30 CS fundamentals MCQ to be solved in 30 mins. Questions are from Database, Computer Networks, C I/O questions, OOP concepts.  

3. Section 3 (Coding Round): This round held for 1 hour and 3 coding questions

   1. Program for Decimal to Binary Conversion
   2. Longest Palindromic Substring | Set 1
   3. Sort elements by frequency | Set 1

4. Section 4: Machine learning basic MCQ questions (10 min 10 questions)

Round 2: It was a normal group discussion

Round 3 (Problem Statement): You are building a financial software system that is intended to be used by millions of ordinary consumers and concurrently by thousands of people. Because it’s dealing with money, there can be no errors in its operation.

1. What steps do you take at different stages of the project to ensure a smooth release and operation of the production?
2. How does your approach changes as the number of concurrent users enter into the millions?

The discussion started, and we have to add to the different stages that are missed by the fellow speaker, concurrency control protocols, transaction, and testing methods, distributed system.  

Round 4 (Interview): Due to covid-19, the interview was on Google meet, and we were provided the link a day before.

• The interviewer asked me to explain the recent project and some cross-questions. Mostly the project was discussed and asked about the inbuilt module I used, algorithm and what are the results, and how I verified them.
• He also asked questions about Machine learning. One of the panel members is working on Machine Learning and one of my projects is also from Machine Learning so then my interview changed track from projects to Machine learning:
  1. What is normalization and why we use it?
  2. Bias variance tradeoff, how to avoid overfitting, Regularization? 
  3. How to measure the accuracy of our model?
  4. Explain your favorite classification technique. 
  5. K means algorithm and what if our dataset is imbalanced like 90% data points belong to one category and the other 10% to the other, which machine learning technique you will use. Support Vector Machine.