Autocoder

Persistent Systems Interview Questions for Freshers

Last updated 2026-11-19

Persistent is a software engineering company, and interviews at product-leaning engineering firms usually go deeper on data structures and computer science than pure support roles. These twelve worked questions cover that depth, with a short design question and HR.

For process basics read how to crack the technical interview; for coding practice use Persistent coding round questions.

RoundsNot published; test, technical and HR commonly reported
DepthStronger DSA and CS fundamentals (typical of product engineering)
CareersCompany's own careers site

Persistent rounds as commonly reported

Persistent does not publish the details of its fresher selection process in a source we could cite. Prep material for engineering-focused firms typically describes an online test with aptitude and coding, a technical interview with stronger data-structures and computer-science fundamentals than at pure services firms, and an HR conversation. Treat the order and difficulty as variable and rely on your invitation for facts.

Where product-engineering interviews go deeper

Hiring status

This page does not track Persistent drives or dates. For anything time-sensitive, use Persistent's own careers site (the careers section of persistent.com) and treat any date quoted on social media or listing sites as unconfirmed until it appears there.

No hiring figures are quoted on this page because there is no sourced statement to cite.

Persistent Systems Interview Questions for Freshers: questions with worked answers

  1. 1

    Why Persistent, and why product engineering work?

    easy

    Persistent is a software engineering company, so show you understand the difference from pure support work: building features, owning quality and working closely with product teams. Give a concrete signal, such as an open-source contribution, a project with real users or a side project with tests and deployment. Keep claims about the company modest and tied to what is on its own website.

  2. 2

    State the time complexity of the main operations of HashMap, TreeMap and PriorityQueue.

    medium

    HashMap: get, put and remove are O(1) on average and O(log n) worst case after treeification in Java 8 or later. TreeMap is a red-black tree, so get, put, remove and ceiling or floor lookups are O(log n), and iteration is in sorted key order. PriorityQueue is a binary heap: offer and poll are O(log n), peek is O(1), and removing an arbitrary element is O(n). Choose TreeMap when you need ordering, PriorityQueue when you need repeated min or max.

  3. 3

    Heap versus binary search tree: when to use which?

    medium

    A heap only guarantees the parent is smaller (min-heap) or larger than its children, so it gives O(1) access to the min or max and O(log n) insert and extract, but searching an arbitrary value is O(n). A BST keeps full ordering, so search, insert, delete, predecessor and range queries are O(log n) when balanced, but it uses more memory and pointer chasing. Use a heap for scheduling and top-K; use a balanced BST (TreeMap) for ordered lookups and ranges.

  4. 4

    Mutex versus semaphore.

    medium

    A mutex is a lock owned by one thread: only the thread that locked it should unlock it, and it enforces mutual exclusion over a critical section. A counting semaphore holds a number of permits; acquire decrements and blocks at zero, release increments, and any thread may release. Use a mutex to protect shared data and a semaphore to limit access to N resources, such as a pool of 10 connections, or to signal between threads. A binary semaphore is similar to a mutex but lacks ownership.

  5. 5

    Three processes arrive at time 0 with burst times 24, 3 and 3. Compute the average waiting time under FCFS and SJF.

    medium

    FCFS in the order P1, P2, P3: waiting times are 0, 24 and 27, so the average is 51 / 3 = 17. SJF runs the shortest first: P2, P3, P1, with waiting times 0, 3 and 6, so the average is 9 / 3 = 3. SJF minimises average waiting time but needs burst estimates and can starve long jobs, while FCFS suffers from the convoy effect. Round robin trades average waiting for responsiveness.

  6. 6

    What is Belady's anomaly?

    hard

    With FIFO page replacement, giving a process more frames can sometimes increase the number of page faults. The reference string 1 2 3 4 1 2 5 1 2 3 4 5 gives 9 faults with three frames and 10 with four. Stack algorithms such as LRU and optimal never show it, because the set of pages held with n frames is always a subset of those held with n + 1. This is a standard check that you understand replacement policies beyond definitions.

  7. 7

    Find the Nth highest salary using OFFSET and using a correlated subquery.

    medium

    With DISTINCT, ORDER BY descending and OFFSET N-1 you get the Nth distinct value. The correlated version counts how many distinct salaries are higher than each row's salary and keeps the row where that count equals N-1. The window-function form with DENSE_RANK is usually clearest. Mention the NULL result when fewer than N distinct salaries exist.

    sql
    SELECT DISTINCT salary
    FROM employee
    ORDER BY salary DESC
    LIMIT 1 OFFSET 2;   -- 3rd highest
    
    SELECT DISTINCT e1.salary
    FROM employee e1
    WHERE 2 = (SELECT COUNT(DISTINCT e2.salary) FROM employee e2 WHERE e2.salary > e1.salary);
  8. 8

    Normalisation versus denormalisation: what is the trade-off?

    easy

    Normalising removes redundancy and update anomalies by splitting data into related tables, which keeps writes consistent but needs joins on reads. Denormalising deliberately duplicates data, for example storing a customer name on each order, to make reads faster at the cost of extra storage and the risk of inconsistency, which you must manage in code. OLTP systems lean towards normalised designs and reporting or read-heavy systems towards denormalised ones.

  9. 9

    How does Python manage memory?

    medium

    CPython uses reference counting: each object tracks how many references point to it and is freed immediately when the count reaches zero. Reference counting cannot free cycles, such as two objects referring to each other, so a cyclic garbage collector runs periodically, organised in generations, to find and free them. Memory is allocated from private pools by pymalloc for small objects. The gc module and weak references give you control, and del removes a name, not necessarily the object.

  10. 10

    Outline a design for a URL shortener.

    hard

    Requirements: create a short code for a long URL, redirect quickly, scale reads. API: POST /shorten and GET /{code} returning a 301 or 302. Generate codes by base62-encoding a unique counter or ID (avoids collisions), or by hashing and handling collisions. Store code to URL in a key-value store or a table indexed on code; put a cache such as Redis in front for hot redirects, since reads far outnumber writes. Discuss expiry, analytics through asynchronous events, rate limiting and horizontal scaling with sharding by code.

  11. 11

    REST versus GraphQL: how would you choose?

    medium

    REST exposes resources at multiple endpoints, benefits from HTTP caching and simple tooling, but clients can over-fetch or need several calls. GraphQL exposes one endpoint where the client specifies exactly the fields it needs, which reduces round trips for varied clients, but caching is harder, queries can be expensive and need depth and cost limits, and the server needs resolvers carefully designed to avoid N+1 queries. Choose REST for simple, cache-friendly APIs and GraphQL for complex, client-driven data needs.

  12. 12

    What motivates you to work on a product team instead of a services-only role?

    easy

    Answer with substance: you like seeing a feature through design, build, release and feedback, you care about maintainability and tests because you will live with the code, and you enjoy talking to users or product managers. Back it with an example such as a student app you maintained for classmates. Do not belittle services work, since many product engineering teams are delivered through services models.

Frequently asked questions

Is Persistent harder than other services companies?

Prep material suggests more emphasis on DSA and fundamentals, but it varies by role and cycle.

Do I need system design?

For fresher roles a simple design question is possible; deep design matters for experienced hires.

What language should I use for coding?

Whichever you know best; Java and Python are typical choices.

Where are Persistent openings?

On the company's own careers site.

Related guides

More in Company Interview Questions

See all in this section