535 questions from 87 interviews · updated from reports 2017-2024
About
CodeVita is a programming contest run by Tata Consultancy Services, where students solve coding problems online and may enter TCS hiring processes. In India, it is linked with entry-level software hiring such as Software Engineer roles and TCS Ninja or TCS Digital profiles.
The roles that come up most are Software Engineer, TCS Digital and TCS Ninja. This covers 87 candidate interviews reported from 2017 to 2024. Most sat it at entry level (86 of 86 that recorded a level). Among the 80 that recorded either route, arrivals split between campus drives (6, 8%) and off-campus applications (74, 92%). Most questions fall under CS fundamentals and DSA.
Ans. C is a procedural, compiled, low-level language with manual memory management, while Java is object-oriented, runs on a virtual machine, and uses garbage collection. C gives more control over memory and hardware, so it is common in systems programming. Java favours portability, safety, and large application development through its standard runtime.
Ans. C++ is a compiled, statically typed language focused on performance and control, while Python is an interpreted, dynamically typed language focused on simplicity and fast development. The most important difference is that C++ gives manual control over memory and hardware-level details, whereas Python manages memory automatically but is usually slower.
Ans. In Python 2, range creates a full list in memory, while xrange returns a lazy sequence object that generates values as needed. The key practical difference is memory use, especially for large ranges. In Python 3, range behaves like Python 2’s xrange, and xrange no longer exists.
Ans. Swap them using arithmetic: set the first number to the sum of both, set the second to the new first minus the second, then set the first to the new first minus the new second. This uses constant space and constant time. The key caveat is integer overflow, so a temporary variable is usually safer.
Ans. A friend function in C++ is a non-member function that is allowed to access the private and protected members of a class. It is declared inside the class using the friend keyword, but it is defined and called like a normal function. It is commonly used for operator overloading or related helper functions.
Ans. Monkey patching in Python is changing or replacing a module, class, method, or attribute at runtime. It works because Python objects are dynamic, so you can assign new behaviour after import. It is often used in tests or quick fixes, but it can make code harder to understand and maintain.
Ans. Hadoop is an open source framework for storing and processing very large datasets across clusters of commodity machines. Its core components are HDFS for distributed storage, YARN for cluster resource management and job scheduling, MapReduce for parallel batch processing, and Hadoop Common, which provides shared libraries and utilities.
Ans. To insert a node into a binary search tree, start at the root and compare the new value with the current node. Go left if it is smaller, right if it is larger, until you find an empty child position. Attach the new node there. Time complexity is O(h), where h is tree height.
Ans. In C, record a timestamp before and after the code using clock_gettime or clock, and in Python use time.perf_counter or the timeit module. The important detail is choosing the right clock: perf_counter and clock_gettime measure elapsed wall time, while CPU-time functions exclude time spent waiting.
Ans. Use a stack to hold operators and build the postfix expression by scanning the infix expression from left to right. Output operands immediately. For an operator, pop higher or equal precedence operators from the stack before pushing it. Push left brackets, and pop until the matching left bracket on a right bracket. This runs in linear time.
Ans. The Internet of Things is a network of physical objects embedded with sensors, software and connectivity so they can collect, exchange and act on data over the internet. These objects can include devices, vehicles, appliances and industrial machines. The key idea is enabling monitoring, automation and control without constant human involvement.
Ans. Python is dynamically typed, usually interpreted, and designed for concise, flexible code, while Java is statically typed, compiled to JVM bytecode, and designed for stronger compile-time checks and large structured applications. The most important practical difference is that Java catches many type errors before running, while Python often finds them at runtime.
Ans. A lambda function in Python is a small anonymous function defined with the lambda keyword. It can take any number of arguments but contains only one expression, whose value is returned automatically. It is commonly used for short callbacks, such as sorting with a key function, where defining a full function would be unnecessary.
Ans. Joins combine rows from two tables based on related columns. Inner join returns only matching rows. Left, right and full outer joins keep unmatched rows from one or both sides with nulls. Cross join returns all combinations. Self join joins a table to itself. Natural or equi joins match columns with equal values.
Ans. Object-Oriented Programming is a programming style that organises software around objects, which combine data and behaviour. Objects are usually created from classes. The key idea is encapsulation: keeping state and the operations on that state together, with controlled access. OOP also commonly uses abstraction, inheritance and polymorphism.
Ans. DFS explores as far as possible along one path before backtracking, while BFS explores all neighbours level by level. DFS uses recursion or a stack; BFS uses a queue. Both mark visited nodes to avoid repeats. For a graph with V vertices and E edges, both run in O(V + E) time.
Ans. TCP is connection-oriented and reliable, while UDP is connectionless and faster but does not guarantee delivery. TCP orders packets, retransmits lost data, and provides flow and congestion control. UDP sends datagrams with minimal overhead, so it is useful for real-time traffic like video calls, gaming, DNS, or streaming where some loss is acceptable.
Ans. Polymorphism is the ability to treat different object types through the same interface while each type provides its own behaviour. For example, different shapes can all have an area method, but each calculates it differently. The key benefit is writing flexible code that depends on common behaviour rather than specific concrete classes.
Ans. Encapsulation is the practice of keeping an object’s data and the methods that operate on it together, while hiding internal details from outside code. The key point is controlled access: fields are usually private, and other code interacts through public methods, which helps protect invariants and reduce unintended dependencies.
Ans. Python has built-in high-level types such as lists, tuples, dictionaries, sets, strings, and arbitrary-size integers, which C does not provide as native data types. C mainly has primitive numeric types, characters, arrays, pointers, structs, and unions, so those Python structures must be implemented manually or via libraries.
Ans. Use printf as the condition of an if statement, with your name as the string argument. The printf call executes while evaluating the condition, so the name is printed, and no semicolon is needed because the if block itself can be empty or contain only braces. This relies on printf returning the number of characters printed.
Ans. A compiler translates the whole program into machine code or another target form before execution, while an interpreter reads and executes the program step by step at runtime. Compiled programs usually run faster after compilation, while interpreted programs are often easier to test and debug because errors appear as the code is executed.
Ans. Arrays in C are fixed-size, contiguous blocks of elements of the same type, while Python lists are dynamic, resizable containers holding references to objects. The key difference is abstraction: C arrays expose low-level memory and require manual size management, whereas Python lists manage resizing and can hold mixed types.
Ans. Use a sieve of Eratosthenes up to the range end, then print numbers marked prime within the requested bounds. Create a boolean array, mark 0 and 1 non-prime, and for each prime p mark multiples from p squared. The time complexity is O(n log log n) and space is O(n).
Ans. An abstract class is a base class that can share state and implemented behaviour, while an interface defines a contract that classes agree to implement. The key detail is inheritance: a class usually extends one abstract class, but can implement multiple interfaces, making interfaces better for capabilities across unrelated classes.
Ans. Artificial Intelligence is the broader field of making computers perform tasks that seem intelligent, while Machine Learning is a subset of AI where systems learn patterns from data instead of being explicitly programmed. The key difference is that AI describes the goal, and Machine Learning is one common method used to achieve it.
Ans. Heap Sort sorts by first building a max heap from the array, then repeatedly moving the largest element at the root to the end and heapifying the remaining part. The key detail is that heapify restores the heap property after each swap. It runs in O(n log n) time and sorts in place.
Ans. Heap memory in Python is the area where objects and data structures are dynamically allocated at runtime. Variables hold references to these objects, not the objects themselves. Python’s memory manager handles allocation and deallocation, mainly using reference counting and garbage collection to reclaim objects that are no longer reachable.
Ans. Python manages memory automatically using a private heap, with objects allocated and freed by the interpreter rather than the programmer. The key mechanism is reference counting, where objects are deallocated when their reference count reaches zero, supported by a cyclic garbage collector to clean up reference cycles that counting alone cannot handle.
Ans. I have used EC2, S3, RDS, Lambda, API Gateway, IAM, CloudWatch, SQS, SNS, Route 53 and CloudFormation. The most important part was using IAM carefully, with least privilege access, and combining CloudWatch logs and metrics to debug production issues and monitor reliability.
Ans. Object parsing in Python means reading structured input and converting it into Python objects such as dictionaries, lists, classes, or dataclasses. The key detail is choosing the right parser for the format, such as json for JSON or xml libraries for XML, then validating the result before using it to avoid type and data errors.
Ans. You make garbage in Java by creating objects and then removing all reachable references to them. For example, reassign a variable, set it to null, or let a local variable go out of scope. The object is then eligible for garbage collection, though the JVM decides when collection actually happens.
Ans. An Adapter Class in Java is a class that provides empty implementations of all methods in an interface, so a subclass can override only the methods it needs. It is commonly used with event listener interfaces in AWT and Swing, reducing boilerplate when an interface has multiple methods.
Ans. Direct access is a file access method where a record can be read or written directly using its address, key, or relative record number, without scanning earlier records. The important point is that storage locations are addressable, so access is much faster than sequential access for known records.
Ans. Bellman-Ford computes single-source shortest paths by initialising all distances to infinity, setting the source to zero, then relaxing every edge V minus 1 times. Store the graph as an edge list of u, v, weight. Finally, scan edges once more to detect a negative-weight cycle. Time complexity is O(VE).
Ans. A VPC, or Virtual Private Cloud, is a logically isolated network in AWS where you run resources like EC2 instances, databases and load balancers. You control its IP address range, subnets, routing, security groups and network ACLs. The key point is that it lets you design secure public and private network areas in the cloud.
Ans. Garbage collection in Java is automatic memory management that finds objects no longer reachable by the program and reclaims their heap memory. The key point is reachability from roots such as stack variables, static fields and active threads. It reduces manual memory errors, but collection timing is not deterministic and may briefly pause execution.
Ans. Self-referential pointers are pointers inside a data structure that point to another object of the same type. For example, a node in a linked list has a pointer to the next node. They are essential for dynamic structures such as linked lists, trees and graphs, where objects are connected at runtime.
Ans. Shell sort repeatedly sorts elements far apart, then reduces the gap until it becomes 1. Start with a gap such as n/2, compare and insert elements within each gap-based sublist using insertion sort, then halve the gap. When the final gap is 1, the array is fully insertion sorted. Average time depends on gaps.
Ans. Python manages memory automatically using a private heap, reference counting, and a garbage collector for reference cycles. Every object has a reference count, and when it reaches zero the object can be freed. The key detail is that del removes a reference, not necessarily the object or memory immediately.
Ans. MongoDB is a document-oriented NoSQL database, while MySQL is a relational SQL database. MongoDB stores flexible JSON-like documents and is good when data shape changes often. MySQL stores data in tables with fixed schemas, joins, and strong relational constraints, which suits structured data and transactions.
Ans. volatile tells the C compiler that an object’s value may change in ways it cannot see, so it must not optimise away or cache accesses to it. Each read or write must be performed as written. It is mainly used for memory-mapped hardware registers and signal-shared variables. It does not make operations atomic or thread-safe.
Ans. Database triggers are stored procedures that run automatically when events such as insert, update or delete occur, while cursors let a program process query results row by row. Triggers are useful for enforcing rules or auditing changes, but can hide side effects. Cursors are flexible, but often slower than set-based SQL operations.
Ans. Use an INNER JOIN by selecting columns from the first table, joining the second table, and matching related keys in the ON condition, such as customer id in both tables. The important detail is that it returns only rows where the join condition matches in both tables, excluding unmatched rows.
Ans. 1NF removes repeating groups by making each field atomic, 2NF removes partial dependency on part of a composite key, and 3NF removes transitive dependency between non-key fields. In practice, each higher normal form assumes the previous one and reduces duplication and update anomalies by ensuring facts depend on the key, the whole key, and nothing but the key.
Ans. The Java Collection Framework is a standard set of interfaces and classes for storing, accessing and manipulating groups of objects. Its core interfaces include List, Set, Queue and Map, with implementations such as ArrayList, HashSet, PriorityQueue and HashMap. The main benefit is consistent APIs, reusable algorithms and predictable performance choices.
Ans. The four pillars of Business 4.0 are intelligence, agility, automation, and cloud. In practice, this means using data and AI for smarter decisions, agile methods for faster change, automation for efficiency and scale, and cloud platforms for flexible, resilient, and cost-effective digital operations.
Ans. Pickling is serialising a Python object into a byte stream, and unpickling is converting that byte stream back into the original object. It is done with the pickle module, often for saving objects to files. The key caution is never unpickle data from untrusted sources, because it can execute code.
Ans. Model the system with Elevator, Request, Scheduler and Controller classes, where each elevator tracks current floor, direction, state and assigned stops. Use two priority queues per elevator for up and down requests, serving nearest stops in the current direction first. New requests are assigned to the cheapest elevator by distance and direction. Operations are logarithmic in queued stops.
Ans. Mark one side of the square into seven equal lengths. From each mark, draw a straight line parallel to the opposite side across the square. This creates seven congruent rectangles. Each rectangle has the same width and the full height of the square, so each has one seventh of the square’s area.
Ans. A strong answer should take a clear, humane stand against caste discrimination while staying respectful and balanced. Emphasise equality, dignity, constitutional values, and social inclusion. Pick examples from education, workplace, or community life where fairness matters. Interviewers listen for maturity, empathy, awareness of social realities, and rejection of prejudice.
Ans. Choose a real dispute where you helped reduce tension and reach a practical outcome. Emphasise listening to both sides, separating facts from assumptions, focusing on shared goals, and agreeing clear next steps. Interviewers listen for calm judgement, fairness, communication skills, and evidence that you protect team relationships while still solving the issue.
Ans. The odd numbers divisible by 3 are 3, 9, 15, 21, and so on. They form an arithmetic sequence with first term 3 and common difference 6. The sum of the first n such numbers is n/2[2(3) + (n − 1)6] = 3n².
Ans. Use Ohm’s law and resistor combination rules. First reduce any series and parallel resistors to an equivalent resistance. Find the total current from the supply using I = V/R. Then work backwards: current is the same in series branches, while voltage is the same across parallel branches. Apply I = V/R to each branch.
Ans. Choose a realistic workplace scenario involving conflict, pressure, ambiguity, or a difficult stakeholder. Emphasise calm judgement, communication, ownership, and respect for process. Show how you would gather facts, involve the right people, act ethically, and protect outcomes. Interviewers listen for maturity, self-awareness, collaboration, and practical problem solving rather than blame.
Ans. Fill the 5-litre mug and pour into the 3-litre mug until it is full, leaving 2 litres in the 5-litre mug. Empty the 3-litre mug. Pour the 2 litres into it. Fill the 5-litre mug again, then pour into the 3-litre mug until full. Exactly 4 litres remain.
Ans. I would propose self-auditing ML models that continuously generate, test and publish their own falsifiable assumptions during deployment. The key detail is a built-in evaluator that compares live outcomes with these assumptions, flags drift or bias, and automatically reduces confidence or asks for human review before harmful decisions scale.
Ans. Make two straight cuts across the top through the centre, at right angles, creating four equal quarters. Then make one horizontal cut through the middle of the cake’s height. Each quarter is split into a top and bottom half, so 4 times 2 gives 8 equal pieces.
Ans. Parse the input date, then add the eclipse cycle assumed by the problem, commonly the Saros cycle: 18 years, 11 days and about 8 hours. Handle leap years and month lengths while adding days. If only a date is required, ignore the time part or round as instructed. Finally print as yyyy-mm-dd.
Ans. Start by clarifying requirements, scale, users, latency, availability, and data consistency needs. Then propose a high level architecture with clients, load balancer, stateless services, database, cache, queue, and monitoring. The most important detail is choosing storage and consistency correctly, because it drives performance, failure handling, and future scaling.
No question matches those filters. Reset filters.
Showing 60 of 535 questions. Ranked by how often the same question came back across interviews.
A spoken interview built from these questions, scored when you finish; the feedback is yours.
Start practisingWhen you are ready, record The One: a single interview hiring teams watch, so you stop repeating first rounds.
Candidate interviews most often cover CS fundamentals (69%) and DSA (21%).
Candidate interviews show an average of 2.8 rounds per experience, with a typical sequence of Online test → Technical → HR. Individual interview paths can vary.
Among questions with a recorded difficulty, the mix is easy 65%, medium 33%, hard 2%.
