Snapdeal Interview Experience | Set 8 (On Campus)

Snapdeal conducted placement drive at my campus in January last week for Software Developer. Eligibility- All CSE (no pointer criteria)

Online Test-

21 (MCQ) +2 (Coding) in 1 hr. Test conducted on hackerrank

21 MCQ had almost 10 aptitude and 11 C output based questions.
Give preference to coding questions. Try to solve both questions ( Pass all test cases for one of the questions and do attempt the other question ( even brute force would pass many test cases )
Aptitude can’t be solved just within a minute. Solve C o/p based questions first.

Questions-
1. Overlapping paintings, find no. of paintings that can be seen distinctly, extreme co-ordinates of paintings are given. Ordering of paintings matter. ( Assume heights of all paintings are same, start and end coordinates are given )

E.g.
5
1 4
2 6
3 4
8 10
7 10

XXXX
   XXXXXX
      XX
           XXX   <- This painting is hidden completely
        XXXX 

Simple O(N^2) solution. Starting from rightmost painting, check if it completely hides any painting or not based on start and end coordinates. ( modification of interval selection problem )

2- Given points of two lines segments A(x1,y1 x2,y2) & B(x3,y3 x4,y4) find whether the 2 segments intersect or not.
Simpler approach ( short code )-
http://community.topcoder.com/tc?module=Static&d1=tutorials&d2=geometry2#line_line_intersection

Length / complicated soln-
https://www.neveropen.co.uk/check-if-two-given-line-segments-intersect/

Expected cut-off –
I solved 2nd question and passed 1 test case for first question ( misunderstood the question during online round !! :p ) and solved 4 MCQ’s only ( all fluke )
So my advice, do solve both coding questions for sure and solve C o/p questions in last 15 min

Round 1 – F2F Technical
Avg 20-30 mins. 22 shortlisted
My went on for 1 hr to 1hr 15 mins

Internship based dicussion (20-30 mins ). Based on Cloud, Virtualization, Networking
Q1- Given N, find LCM from of all numbers from 2 to N. Give the complexity expressed in the form of Number of prime numbers <= N. Had to be really precise in terms of complexity ( in terms of prime factors, maximum recurrences, each recurrence complexity ). Long dicussion on complexity. Don’t say any method whose complexity you cannot prove. (E.g saying that i can use Sieve of Eratosthenes for prime pre-processing will lead to question of complexity of Sieve which is O( Log LogN), that cannot be proved trivally. ) So avoid using any such terms

Q3- Spring / Hibernate in JAVA
Told him i work in C/C++ only. No experience in JAVA

Q2- Types of SQL- NoSQL and SQL(Relational DBMS ). Why the need of NoSQL- Big Data Analytics

Q3- How would you design DBMS for Snapdeal’s website’s shoe section.
Now if you want to further break it into Sports and Casual Shoe would you break the DB into two or add another entity ? Full justificdation
I initially answered with a multi-level indexed structure for DBMS storage. Could not answer on the second part of the question. He asked if i knew DBMS and i told him I do not know DBMS. He skipped the question and ended the interview. Told him i had advanced DBMS lab in my course currenlty and would learn it before graduating.

Round 2- Coding Round ( 2 Hrs )
10 shortlisted
Q1- Turn an image by 90 degree
Q2- Given a sequence of words, print all anagrams together
Q3- Find a triplet that sum to a given value

I was the first one to solve all 3 in 45 mins roughly and went for next interview. Shortlisting criteria- 2 questions in 1 hr – 1hr 15 mins even though they said that we had 2 hrs to solve all 3 questions !!

Round 3- F2F Technical
4 shortlisted. This round went for almost 1hr 45 min – 2 hrs for me since I solved the Round 2 question earliest. Other 3 had almost 45 mins interview.

Q1- Variation of
Print all possible words from phone digits
Given a dictionary of words and a number n. Find count of all words in dictionary that can be formed by given number n.
I started by exponential solution and reduced it to polynomial. We discussed various approaches and tried a variety of methods and after 1-1.5 hrs of discussion finally ended up with an O(1) solution with some pre-processing overhead. After achiveing O(1) time complexity, he asked to further optimize the space complexity.
Usage of Trie / TST. Internal Implementation of Hashing structure and replacing the hashing mechanism using Trie / TST.

Q2- Given an array of elements. We can perform following operation only- Increase an array element. Cost of operation is the amount of increment made per array element. Now for a given H, we need to make any H ( not necessarily consecutive ) elements of array equal with minimum cost.

E.g.
N=6, H=4
2 3 5 6 4 4

changes to -> 4 4 5 6 4 4
Cost is ( 4-2 + 4-3 = 3 )

N=6, H=3
2 3 5 6 4 4

changes to -> 2 4 5 6 4 4
Cost is ( 4-3 = 1 )
Optimal complexity- O(N)

Round 4 -F2F (HR)
3 shortlisted

Typical HR round.

I would like to thank neveropen for a exhaustive set of interview questions and study material on data structures-algorithms.

If you like neveropen and would like to contribute, you can also write an article and mail your article to review-team@neveropen.co.uk. See your article appearing on the neveropen main page and help other Geeks.