90 questions from 15 interviews · updated from reports 2019-2025
About
KPIT Technologies is an Indian technology company providing software and engineering services for the automotive industry, including embedded systems, electrification, connected vehicles, and autonomous driving. In India, it hires Associate Engineers, Associate Software Engineers, and Software Engineers for development, testing, and embedded software roles.
The roles that come up most are Associate Engineer, Software Engineer and ASE (Associate Software Engineer). This covers 15 candidate interviews reported from 2019 to 2025. Most sat it at entry level (15 of 15 that recorded a level). Among the 13 that recorded either route, arrivals split between campus drives (11, 85%) and off-campus applications (2, 15%). Most questions fall under CS fundamentals and DSA.
Ans. An lvalue is an expression that refers to a persistent object or memory location, while an rvalue is a temporary value that does not persist beyond the expression. Typically, lvalues can appear on the left of assignment. The key practical detail is references: lvalue references bind to lvalues, and rvalue references bind to temporaries, enabling move semantics.
Ans. OOP in C++ is implemented mainly through classes and objects. Classes group data and functions, while access specifiers provide encapsulation. Inheritance lets one class reuse or extend another. Polymorphism is achieved with function overloading, overriding, and virtual functions. Abstraction is commonly done using abstract classes with pure virtual functions.
Ans. Create a new process in Linux by calling fork, which duplicates the current process into a parent and child. The key detail is that fork returns different values: zero in the child, the child’s PID in the parent, and negative on failure. The child often calls exec to run a new program.
Ans. Choose the search method based on the data: use linear search for unsorted data, binary search for sorted data, and BFS or DFS for graph or tree searches. The key detail is recognising whether the search space can be reduced each step, which gives binary search O(log n) time instead of O(n).
Ans. Memory management in C/C++ means controlling how memory is allocated, used, and released, and pointers are variables that store memory addresses. Stack memory is automatic and scoped, while heap memory is explicitly allocated and must be freed. The key risk is invalid pointer use, causing leaks, dangling pointers, buffer overflows, or undefined behaviour.
Ans. Implement responsive design by using a mobile-first CSS layout with flexible grids, relative units, responsive images, and media queries to adjust styles at different viewport widths. The most important detail is to design for content first, then add breakpoints where the layout naturally needs to change, not just for specific devices.
Ans. A diode is a two-terminal component that mainly lets current flow in one direction, while a transistor is a three-terminal component used to switch or amplify signals. The key difference is control: a diode responds to applied voltage, but a transistor uses one terminal to control current between the other two.
Ans. Props are read-only values passed from a parent component, while state is data owned and managed inside a component. Props let components receive configuration or data. State represents values that can change over time, such as user input or loading status, and must be updated through React’s state update functions to trigger re-rendering.
Ans. The four pillars of OOP are encapsulation, abstraction, inheritance, and polymorphism. A typical demonstration uses a base class such as Animal with a speak method, subclasses like Dog and Cat overriding it, private fields accessed through methods, and an abstract interface hiding details. Time complexity is not the focus here.
Ans. Sorting algorithms arrange data for easier processing, while searching algorithms find required items efficiently. Quick sort and merge sort are common for general sorting, heap sort suits priority-based work, and counting sort works for limited integer ranges. Linear search fits unsorted small data, while binary search is best on sorted data, with logarithmic time.
Ans. A strong answer should pick a real mistake with manageable impact, then show ownership, calm judgement, and escalation. Emphasise checking facts, informing the right available person, fixing what you can, documenting decisions, and updating the lead on return. Interviewers listen for accountability, not blame or panic.
Ans. Implement common sorts by comparing and swapping or partitioning elements: bubble, selection and insertion sort are simple in-place methods with O(n²) time, while merge sort and quicksort use divide and conquer. Merge sort is O(n log n) with extra space. Quicksort averages O(n log n), but worst case is O(n²).
Ans. Choose a real, low-risk missed deadline where you took ownership rather than blamed others. Emphasise when you realised the risk, how quickly you communicated, what trade-offs or support you proposed, and how you reduced impact. Interviewers listen for accountability, calm problem solving, stakeholder management, and clear learning that changed your future planning.
Ans. Polymorphism is mainly compile-time polymorphism, through method overloading or operator overloading, and runtime polymorphism, through method overriding. For runtime polymorphism, create a base class with a virtual or overridable method, override it in child classes, and call it through a base reference. It uses a virtual table, with O(1) dispatch.
Ans. Pick a real but low-risk example where personal preference, loyalty, or outside relationships could affect judgement. Emphasise that you recognised it early, disclosed it to the right person, followed company policy, and removed yourself from biased decisions if needed. Interviewers listen for integrity, transparency, sound judgement, and respect for governance.
Ans. A microcontroller is a complete small computer on one chip, while a microprocessor is mainly just the CPU. A microcontroller usually includes memory, timers and input/output peripherals, so it can control embedded devices directly. A microprocessor normally needs external memory and peripherals, but offers more processing power and flexibility.
Ans. Use a quick structure: identify the rule, ignore distracting detail, test the smallest possible cases, then eliminate impossible options. For sequences, look for differences, ratios, positions, or alternating patterns. For arrangements, draw a simple table or line. If stuck after a minute, mark it, move on, and return later.
Ans. Choose a real work problem with clear stakes, where your actions changed the outcome. Emphasise how you analysed the issue, involved others, weighed options, and followed through. Interviewers listen for ownership, calm judgement, practical problem solving, communication, and learning, not a perfect result or blame placed elsewhere.
Ans. Choose a real example where the disagreement affected delivery, not personal preference. Emphasise listening first, clarifying goals, comparing evidence, and agreeing decision criteria. Show you stayed respectful, involved others only when needed, and committed once a decision was made. Interviewers listen for maturity, collaboration, and focus on outcomes over ego.
Ans. C storage classes are auto, register, static and extern. auto variables have block scope and automatic storage, usually on the stack. register variables have block scope and may be kept in CPU registers. static variables have static storage in data or BSS, with block or file scope. extern declares a global object stored in data or BSS.
Ans. In C++, class and struct are almost the same, except class members are private by default and struct members are public by default. Use a class for encapsulated objects with behaviour and invariants, a struct for simple data aggregates, and a union when only one of several fields is valid at a time.
Ans. Pointers in C are variables that store memory addresses, usually the address of another object or function. They can be reassigned, set to null, incremented, and used for manual memory and array access. C++ references are aliases for existing objects, must normally be initialised when created, cannot be reseated, and are used with safer syntax.
Ans. Answer conceptual CS questions by starting with the core definition or decision, then add the main trade-off, mechanism, or example. For operating systems, networks, embedded systems, DBMS, and DSA, focus on correctness, resource limits, concurrency, latency, consistency, and time or space complexity where relevant. Keep explanations concise and practical.
Ans. Use bitwise shifts: left shift by one to multiply by 2, and right shift by one to divide by 2. In C, use n << 1 and n >> 1. This is constant time, O(1), and needs no extra data structure. For negative signed values, right shift behaviour can be implementation-defined.
Ans. The CSS box model describes how every element is laid out as a rectangular box made of content, padding, border, and margin. By default, width and height apply to the content only, so padding and border add to the total size. Using box-sizing: border-box makes sizing more predictable.
Ans. A foreign key is a column or set of columns in one table that refers to the primary key or unique key of another table. It links related tables and enforces referential integrity, meaning a row cannot reference a non-existent related row unless the foreign key is allowed to be null.
Ans. File handling in C++ is done using stream classes from the fstream library: ifstream for reading, ofstream for writing, and fstream for both. A file is opened with a mode such as input, output, append, or binary, then data is read or written using stream operators or functions. Always check the stream state for errors.
Ans. Create a form in HTML with a form element containing labelled input controls and a submit button. The form element usually sets an action URL, which receives the submitted data, and a method, usually GET for simple queries or POST for data that changes state or sends sensitive content.
Ans. ACID properties are the guarantees that make database transactions reliable: Atomicity, Consistency, Isolation and Durability. Atomicity means all or nothing, Consistency keeps valid rules, Isolation prevents concurrent transactions interfering, and Durability ensures committed changes survive crashes. They are essential for correctness in systems handling critical data.
Ans. Structures are used in C and C++ to group related variables of different data types under one name. They help model real-world entities, make data easier to manage, and allow functions to pass or return a single compound object instead of many separate values.
Ans. Kill a Linux process by sending it a signal, usually with kill followed by the process ID. First find the PID using ps, top, or pgrep, then send SIGTERM for a clean shutdown. If it does not stop, use SIGKILL with kill -9, but only as a last resort.
Ans. Namespaces in C++ are used to group related names and prevent naming conflicts between variables, functions, classes, and libraries. They define a named scope, so the same identifier can safely exist in different namespaces. For example, the standard library uses std, which keeps its names separate from user-defined names.
Ans. Polymorphism in C++ is the ability of one interface or name to represent different behaviours depending on the object or argument types. Its main types are compile-time polymorphism, using function overloading, operator overloading and templates, and run-time polymorphism, using inheritance with virtual functions called through base class pointers or references.
Ans. Inheritance in C++ lets a class derive properties and behaviour from another class, enabling reuse and polymorphism. The main structural types are single, multiple, multilevel, hierarchical and hybrid inheritance. The important detail is the inheritance access specifier: public, protected or private, which controls how base class members are exposed in the derived class.
Ans. Java is platform independent because Java source code is compiled into bytecode, not directly into machine-specific native code. This bytecode runs on the Java Virtual Machine, which is available for different operating systems and hardware. As long as a compatible JVM exists, the same compiled program can run unchanged.
Ans. C is mainly a procedural systems programming language, while C++ extends C with object oriented and generic programming features. The key practical difference is that C++ provides classes, constructors, destructors, templates and a richer standard library, enabling abstractions such as RAII and containers while still supporting low level memory control.
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. Scan the array once and keep a variable holding the smallest value seen so far. Initialise it with the first element, then compare each remaining element and update it when a smaller value is found. This uses no extra data structure beyond one variable, takes O(n) time and O(1) space.
Ans. Use two pointers and swap characters in place, one starting at the beginning and one at the end. Move both pointers towards the centre after each swap until they meet or cross. This uses the character array itself as the data structure, runs in O(n) time, and uses O(1) extra space.
Ans. C is a procedural language, while C++ is largely a superset of C with object oriented and generic programming features. C++ adds classes, inheritance, polymorphism, templates, exceptions, references, function overloading and the standard library. The most important difference is abstraction: C gives low level control, while C++ supports higher level design without losing that control.
Ans. An enum, or enumeration, is a data type that defines a fixed set of named constant values. It is used when a variable should only take one value from a known list, such as days of the week or order status. Enums improve readability, reduce invalid values, and make code easier to maintain.
Ans. Header files are files, usually in C or C++, that contain declarations such as function prototypes, class definitions, constants, macros and type definitions. They are used to share interfaces between source files without duplicating code. The key point is that headers declare what exists, while source files usually define how it works.
Ans. Loops repeat a block of code, while switch statements choose one block from several alternatives. In C/C++, for, while and do while loops control repetition using conditions or counters. A switch evaluates an integral or enum expression and jumps to a matching case. break prevents fall-through, and default handles unmatched values.
Ans. Repeatedly take the last digit using modulo 10, multiply it into a running product, then remove the digit using integer division by 10. Use the absolute value for negative numbers. If the number is 0, the product is 0. This uses no extra data structure and runs in O(d) time, where d is the digit count.
Ans. Data types define the kind of value a variable can hold and the operations allowed on it. Common types include primitive types such as integer, float, character and boolean, and non-primitive or composite types such as arrays, strings, structures, classes, lists and maps. They help ensure correct storage and processing.
Ans. Lists and tuples are ordered collections in Python, used to store multiple values in a single variable. Lists are mutable, so use them when items need to be added, removed, or changed. Tuples are immutable, so use them for fixed data, safer returns from functions, or values that should not change.
Ans. The main difference is that C is a procedural language, while C++ supports both procedural and object-oriented programming. C++ adds classes, objects, encapsulation, inheritance and polymorphism, which help structure larger programs. C is usually closer to low-level system programming and has a smaller language feature set.
Ans. BFS visits a graph level by level using a queue, while DFS goes as deep as possible before backtracking using recursion or a stack. Both mark visited nodes to avoid cycles. BFS is useful for shortest paths in unweighted graphs. DFS is useful for connectivity, cycle checks, and topological-style exploration. Both run in O(V + E).
Ans. Use the Sieve of Eratosthenes to find all primes less than or equal to n. Create a boolean array marking every number as potentially prime, then for each number from 2 up to √n, mark its multiples as not prime. The remaining marked numbers are primes. Time complexity is O(n log log n), space is O(n).
Ans. Java does not support multiple inheritance of classes, so a class cannot extend more than one class. It can achieve a form of multiple inheritance through interfaces, because a class may implement many interfaces. If default methods conflict, the implementing class must explicitly override and resolve the ambiguity.
Ans. A year is a leap year if it is divisible by 400, or if it is divisible by 4 but not divisible by 100. Check these conditions in that order or as one boolean expression. No data structure is needed, and the time complexity is O(1).
Ans. A stack is a last in, first out structure, while a queue is a first in, first out structure. In C++, use std::stack for push, pop and top, and std::queue for push, pop and front. These operations are usually constant time. A stack suits undo actions; a queue suits task scheduling.
Ans. The JVM runs Java bytecode, the JRE provides the JVM plus standard libraries needed to run Java programs, and the JDK includes the JRE plus development tools such as the compiler and debugger. In practice, users need a JRE to run applications, while developers need a JDK to build them.
Ans. Semantic HTML elements are tags that describe the meaning and structure of their content, not just how it looks. Examples include header, nav, main, section, article, aside, footer, figure and time. They improve accessibility, make pages easier for search engines to understand, and help developers read the document structure.
Ans. Inheritance is implemented by defining a base class with shared state and behaviour, then defining a derived class that extends it and adds or overrides members. The important detail is method dispatch: runtimes usually store class metadata, often a virtual method table, so calling an overridden method is typically O(1).
Ans. An array is usually fixed in size, while a vector is a dynamic array that can grow or shrink. Arrays may be allocated on the stack or heap and have minimal overhead. A vector manages its own storage, supports operations like push and pop, and may reallocate when capacity is exceeded.
Ans. Use nested loops: the outer loop controls rows, and the inner loop prints the required characters or spaces for each column. Derive the print condition from the pattern’s row and column positions. No extra data structure is usually needed. Time complexity is O(rows × columns), with O(1) extra space.
Ans. Computer networks let devices exchange data using layered protocols, mainly the TCP/IP model. The key idea is separation of responsibilities: IP handles addressing and routing, TCP provides reliable ordered delivery, UDP provides faster best-effort delivery, DNS maps names to IP addresses, and application protocols such as HTTP define how services communicate.
Ans. Read the question first, then the passage or sentence with a clear purpose. For grammar, check subject verb agreement, tense, pronouns, modifiers, and parallel structure. For vocabulary, use context clues and word roots. For comprehension, identify the main idea, tone, evidence, and eliminate options that are too broad, too narrow, or unsupported.
Ans. Identify what is being asked, list the given values, and choose the relevant formula or relationship. For algebra, form equations and solve systematically. For geometry, draw a diagram and use standard properties. For arithmetic, simplify step by step. For data interpretation, read labels carefully, calculate only what is needed, and check units.
No question matches those filters. Reset filters.
Showing 60 of 90 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 (61%) and DSA (23%).
Candidate interviews show an average of 2.7 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 73%, medium 27%, hard 0%.
