Nagarro, Persistent, Publicis Sapient, Thoughtworks: How Each Interviews Freshers in India

For anyone with an off-campus link or campus slot from one of these four. They sit between the IT-services pipeline and the product unicorns, and their prep advice online is mostly one-person GeeksforGeeks posts from different years.

TL;DR: Nagarro and Persistent run the online-test-first format you know from service firms, on Mettl and an AMCAT-style platform. Publicis Sapient's format changes between drives, but its last round is about values, and that is where one candidate reported being rejected after clearing the code. Thoughtworks is the outlier, with a pairing session where finishing the problem matters less than how you talk and test. Round counts and question counts vary by drive and year at all four.

I read the candidate posts that are still up, mostly 2023 to 2025, plus the official Thoughtworks pages, and kept only claims I could see in a page. Where the sources disagree, I say so instead of averaging them.

Nagarro: Two Mettl Tests, Then a Short Technical Round

The reported flow is an online test, a coding test, a technical interview and HR. Both a July 2025 post and a February 2023 post put the first round on Mettl with 25 aptitude MCQs and 15 DSA or C++ MCQs. The coding test is where accounts split. A 2025 post lists 3 questions (a longest-increasing-subsequence style DP, a string problem, a tree problem). A 2023 post says 5. An on-campus post says 5 in 60 minutes: three very easy, one medium, one hard, with at least three solved to clear.

So don't prep for "the Nagarro number". Prep to solve the easy three fast and cleanly, because that seems to be what carries people through.

The technical interview is reported as light compared with the test: sorting algorithms, OOP, SQL (joins, second-highest salary, normalisation) and a linked-list or BST question over screen-share. One post has a single interviewer on Google Meet, another says about an hour. HR is short, around 7 to 8 minutes, and asks why Nagarro, where you see yourself in five years, whether you'll relocate, and what other offers you hold. In a 2024 off-campus account the first assessment link came a week after applying and the offer a day or so after HR.

Two pages you may find are stale. One describes a pen-and-paper written round and another a three-hour coding test. Both conflict with the newer posts, so I'd ignore them.

Persistent Systems: The Objective Round Does the Cutting

Persistent's online assessment is reported on an AMCAT-family platform. CCBP's write-up (updated December 2025) describes an objective round of 44 questions in 50 minutes: 20 computer science in 20 minutes, 12 English in 15, 12 logical in 15. After that comes a two-question automata coding round of 45 minutes, and an advanced coding round of two questions in 60 minutes for top scorers. A GeeksforGeeks post about a 2022 campus drive reports 46 questions, an SHL interview app and results five days later, which roughly matches.

Interviews then run in two technical rounds and an HR round. Reports have round one on projects, DBMS and algorithms, and round two on data structures, operating systems and networking. One campus post has an expression-validity question and product-of-array-except-self. HR is described as relaxed: confidence, family background, why should we hire you.

One thing worth planning around: a single campus post says only 8 candidates were shortlisted for interviews after the test, against 60-plus hired the previous year. That is one drive, but it fits a test that filters hard.

On money, there's no official figure. Coaching and aggregator sites say ₹4.5 to 6 LPA for a base fresher and mention the advanced coding round lifting the package into the high ₹7 LPA range. One of those sites says outright that its numbers are "not official salary tables". Use them to compare offers, not to negotiate with.

Publicis Sapient: The Format Moves, the Values Round Doesn't

This is the least stable of the four. A 2023 campus post describes a 100-minute test with 33 MCQs (reasoning, DSA, OOP, RDBMS, C++ or Java) plus two coding questions (decode ways, Two Sum), then an in-office technical round on JavaScript async, Java output prediction and number-to-words, then 25 minutes of HR with situational questions.

A July 2025 campus post has a different shape: two medium string problems on HackerRank in 75 minutes, of which only 12 of about 135 candidates advanced. The domain round ran 50 to 60 minutes on Hirepro with live coding and project questions, followed by an "attribute" interview. Two students were selected. An April 2025 post has a 90-minute assessment with two coding questions, a technical round on tries, priority queues, a library-management design, threads and OS, and a Core Values round with conflict and leadership questions. That author cleared the technical round and was rejected in the last one.

Whatever the platform, both 2025 posts end in a values or attribute conversation, and that's the round to rehearse out loud. Prepare two stories (a disagreement in a team project, a time you led something) with specifics you can stand behind, and practise saying them in under two minutes.

I found no reliable fresher salary figure for Publicis Sapient, so I'm not going to give one.

Thoughtworks India: The Only One Where Pairing Is the Test

Thoughtworks publishes its process. The official page describes a recruiter screen, role-specific interviews (a pairing session or a case), then a leadership and culture round, averaging four to six weeks and mostly remote.

The graduate accounts agree with that shape. A December 2024 campus post has a HackerRank round of 4 questions in 75 minutes (19 of 175 advanced), then code pairing (8 advanced), then an hour of technical questions on trees, linked lists, SQL and projects, then a values round. Two were hired. A 2024 off-campus post has two coding rounds, a pairing repo that arrived 30 minutes before a session of about 1.5 hours, a technical interview near two hours and a cultural round of about an hour, with days to weeks between stages.

The pairing format is the reason this firm needs different prep. Thoughtworks' own pairing write-up says there are two sessions, you aren't expected to finish, and they assess communication, design, teamwork and unit testing. Solving LeetCode problems silently doesn't rehearse that. Pick a small exercise (a shopping cart, a parking lot), set a timer, and practise writing a test first, explaining why, and changing your design when your pair suggests something.

Their careers FAQ also says GenAI use is allowed if you disclose it. That's a Thoughtworks statement, not an industry norm.

On pay, the only figure I found was a student-reported "about ₹10 to 18 LPA" range from one salary-guide site, so I'd treat it as a rough signal only.

Side by Side

How to Split Your Prep

Nagarro or Persistent: these are timed-test firms. Drill the objective sections under a clock, since that's where the cutting happens, and keep easy coding problems fast. For the coding rounds on Mettl, AMCAT-style and HackerRank platforms, OA Copilot (online assessments on HackerRank, CodeSignal, Codility and similar platforms) is built for exactly this format, and you should read each platform's proctoring terms before you decide how to use anything. Check your invite for what the drive allows.

Publicis Sapient: ask which platform and how many coding questions your drive uses, then put your remaining time into the values round.

Thoughtworks: practise talking while you code. A mock interview run with a timer, or a friend on a call, gets you used to explaining decisions on camera, which is the skill the pairing session scores.

For all four: most live rounds are remote. Check your camera, your connection and your screen-share setup the day before, because a dropped call in a 45-minute slot costs more than any single wrong answer. If you're deciding between these and the bigger names, the Capgemini, Accenture and HCLTech comparison and the AMCAT and eLitmus guide cover the platforms that turn up here too.

FAQ

How many rounds does the Nagarro interview have for freshers?

Recent posts describe an online test, a Mettl coding test, one technical interview and HR. The coding test count is what changes: 3 questions in a 2025 post, 5 in a 2023 post, and 5 in 60 minutes in an on-campus post with three needed to clear. Check your own invite.

Is the Persistent Systems test hard?

The objective round is the one people underrate: 44 questions in 50 minutes across computer science, English and logic, per CCBP. Practise it against a timer, since running short there ends the drive before your coding is read.

Does Publicis Sapient hire through AMCAT or HackerRank?

Both appear in different sources. A 2023 campus post has a 100-minute test with MCQs and two coding questions, and a 2025 campus post has two string problems on HackerRank in 75 minutes. Your invite will tell you which one you're getting.

What is a Thoughtworks pairing interview?

Two sessions where you aren't expected to finish. They assess communication, design, teamwork and unit testing, and candidates report getting a repo shortly before it starts. Narrate your choices and write tests as you go.

Can I use AI in a Thoughtworks interview?

Their careers FAQ says GenAI use is allowed if you disclose it. I found no similar statement from Nagarro, Persistent or Publicis Sapient, so follow the rules in your own invite.

What is the fresher salary at Persistent Systems?

There's no official table. Aggregator sites cluster around ₹4.5 to 6 LPA for a base fresher, with one reporting the advanced coding round lifting it into the high ₹7 LPA range. Confirm in your offer letter.

---

Author · Alex Chen. Career consultant and former tech recruiter. Spent 5 years on the hiring side before switching to help candidates instead. Writes about real interview dynamics, not textbook advice.