Project Defence Plus DSA round·Engineering·Medium·20 min

Tata Consultancy Services Digital Trainee Interview — Project Defence Plus DSA

20 min · 1 credit · scorecard at the end

ZeroPitch Research Desk

Updated 2026-05-25 · Newly added · be among the first to take it

Field: Engineering
Company: Tata Consultancy Services
Role: Digital Software Engineer Trainee
Duration: 20 min
Difficulty: Medium
Completions: New
Updated: 2026-05-25

What this round is about

Topic focus. A TCS Digital technical round in India for the 7.5 LPA premium track, covering project deep-dive, one medium DSA problem, URL shortener design lite, and one advanced fundamentals probe on DBMS or SOLID.
Conversation dynamic. The interviewer pushes back twice on every vague answer before moving on, exactly like a real TCS Chennai Industry Senior Member.
What gets tested. Your ability to defend technology choices, dry-run code on paper, name system components and walk the happy path, and connect SOLID or indexes to a real project example.
Round format. 20 minutes, voice plus whiteboard, four blocks: warm-up project defence, DSA coding, URL shortener sketch, advanced fundamentals plus closing why-Digital probe.
Audience anchor. Tuned for SMVEC and similar Tier-2 college NQT 80+ percentile candidates aiming for the Digital tier rather than dropping back to the Ninja band.

What strong answers look like

Module-level project ownership. You name the exact module you personally built (for example I owned the JWT authentication layer or I wrote the appointment-scheduling logic), not what the team built.
Technology trade-off with one alternative. You name the technology you chose, name the alternative you considered, and give one sentence of reasoning for the rejection (for example I picked React over Angular because the tutorial set we used was React-first and our team had three weeks).
Talk-through-then-code on DSA. You state your approach, the data structure, and the time complexity out loud before you write the first line of code. You then dry-run on a three or five element input on paper, finger pointing at each step.
Component coverage on system design lite. You name at least four components for the URL shortener (load balancer, web server, shortener service, database) and walk through the redirect happy path in one breath.
SOLID with code, not definitions. When asked about Single Responsibility or Liskov, you reach for a class from your own project rather than reciting the textbook definition.
Specific why-Digital reason. Your why-Digital answer cites the BFSI client breadth, the seven-and-a-half LPA scope, or the Chennai allocation pattern, not generic ambition.

What weak answers look like (and how to avoid them)

The we-built-it trap. Saying we built the appointment module when asked what you personally built. Fix: switch to I owned X within five seconds of any project probe.
The textbook recital trap. Reciting the OOP four pillars or the SOLID acronym without a code or project anchor. Fix: keep one Java or Python class in your head for every concept and reach for it first.
The memorised code trap. Writing a linked-list cycle detection function from muscle memory but failing to dry-run it on a four-node input on paper. Fix: practise the dry-run, not the keystrokes.
The component-list trap. Naming load balancer, server, database for the URL shortener but skipping the redirect walk-through. Fix: state the request path from click to HTTP 302 in one breath.
The generic why-Digital trap. Answering good company or family pressure or I want to grow professionally. Fix: tie your answer to one specific TCS engagement, the Chennai delivery model, or the BFSI Digital practice.
The resume-inflation trap. Listing Spring Boot or Dijkstra on the resume but stumbling on dependency injection or when-to-prefer-BFS. Fix: every line on your resume gets one rehearsed three-sentence explanation.

Pre-interview checklist (2 minutes before you start)

Recall the one module you personally owned. Have the exact filename, class name, or feature name ready before the round opens.
Have one technology alternative in mind. Know the rejected alternative for every technology on your resume, with one sentence of reasoning.
Think of one SOLID principle you used. Pick the one principle that actually shaped a class in your project and rehearse the two-sentence walk-through.
Pull up the URL shortener component list. Recite load balancer, web server, shortener service, database in your head once before the round.
Re-read the four isolation levels. Have READ UNCOMMITTED, READ COMMITTED, REPEATABLE READ, SERIALIZABLE on the tip of your tongue with the anomaly each prevents.
Identify your specific why-Digital reason. Pick one concrete anchor about TCS Chennai BFSI breadth or the seven-and-a-half LPA scope before you sit down.

How the AI behaves

Probes every claim twice. If you say I used React or I added an index, expect two follow-ups asking why and how before the topic closes.
No mid-interview praise. Aarav will not say great answer or perfect or that is strong. He acknowledges the specific content of what you said, then pushes deeper.
Interrupts on the word we. If you describe a project at team level he will stop you and ask which module you personally built.
Pushes you to the whiteboard early. Within two turns of any visual question he will direct you to pull up the whiteboard and start sketching rather than narrating.
Marks resume inflation calmly. If you cannot defend a technology you listed on your resume he will mark it and revisit at the wrap-up.
Never coaches the answer. He never names the framework you should use, never enumerates the components you should list, never tells you the correct isolation level. He only probes.

Common traps in this type of round

Team-level project description. Using we built throughout a project explanation without ever stating which module you personally owned.
Pre-memorised DSA code without a dry-run. Writing a cycle detection or middle-node function correctly from memory but freezing when asked to walk through with a five-node input.
Component list without happy path. Listing four URL shortener components and stopping there, leaving the redirect flow unexplained.
OOP and SOLID textbook recital. Naming the four pillars or all five SOLID letters but never reaching for a class from your own codebase.
Wrong-answer-by-resume-inflation. Claiming Spring Boot on the resume and stumbling on dependency injection when the interviewer probes.
Generic why-Digital reason. Answering good company or family pressure or I want stable career instead of a specific TCS Chennai or BFSI Digital anchor.

Sample problems you'll face

The 2 problems below are the same ones you'll work through in the live session — no surprises. Read the constraints carefully; the AI persona will refer you to the on-canvas card by problem number.

1Middle of a Linked List
Given the head of a singly linked list, return the node at the middle. For even-length lists return the SECOND of the two middle nodes. Solve in a single pass.
Example input1 → 2 → 3 → 4 → 5
Example outputNode with value 3
- 1 ≤ list length ≤ 10^4
- Single pass — slow/fast pointer (Floyd's tortoise-and-hare)
- Walk through the pointer positions after each iteration on n=5 and n=6 before writing code
- Use the on-canvas card to read the prompt; sketch or write code on the whiteboard. The AI sees what you draw.
2Detect Cycle in a Linked List
Return true if the linked list contains a cycle, false otherwise. Use O(1) extra space.
Example input1 → 2 → 3 → 4 → 2 (cycle back to node with value 2)
Example outputtrue
- 0 ≤ list length ≤ 10^4
- O(1) space — no hash set
- Explain WHY the two pointers must eventually meet if a cycle exists
- Use the on-canvas card to read the prompt; sketch or write code on the whiteboard. The AI sees what you draw.

Interview framework

You will be scored on these 7 dimensions. The full rubric with definitions is below.

Project Ownership Specificity

How precisely you name the module you personally built, the technology you chose, and the alternative you rejected, with one sentence of reasoning per choice.

20%

Dsa Dry-run Rigor

How clearly you state the approach and complexity before writing code, then dry-run on a small sample input on the whiteboard rather than reciting from memory.

20%

URL Shortener Design Coverage

Whether your sketch names at least four components and walks the redirect happy path from click to HTTP response in one breath.

15%

Advanced Fundamentals Application

How well you tie SOLID, design patterns, indexes, or isolation levels to a class from your own project rather than reciting the textbook definition.

15%

Composure Under Pushback

How you handle the interviewer's two-round pushback on every vague answer — revising with detail rather than getting defensive or blaming the team.

10%

Digital Track Motivation Specificity

Whether your why-TCS-Digital answer cites a specific BFSI engagement, Chennai allocation, or 7.5 LPA scope rather than generic ambition or stability talk.

10%

Resume Claim Defensibility

Whether you can defend every technology you listed on your resume with one rehearsed three-sentence explanation when probed.

10%

What we evaluate

Your final scorecard breaks down across these dimensions. The full rubric and tier criteria are revealed inside the interview itself.

Common questions

What does the TCS Digital technical round actually test?

The TCS Digital technical round tests four things across thirty to forty-five minutes. Project ownership under sharp questioning, including technology trade-off justification. One or two medium-difficulty DSA problems on paper or a shared editor covering linked lists, trees, graphs, or dynamic programming basics. System design lite where you identify components for a problem like a URL shortener and walk through the happy path. Advanced computer science fundamentals including indexes, isolation levels, SOLID with code examples, and design patterns. The interviewer pushes back twice on any vague answer before moving on.

How should I structure my project deep-dive answer?

Start with a one-line description of what the project does. Then immediately name the specific module you personally built. Not what the team built, what you built. Name the technology you chose for that module and one alternative you considered and rejected. Be ready to defend the choice in one sentence. The interviewer will push twice. First on the technology rationale. Then on a deeper implementation detail like how authentication works or what indexes you added. Stay specific. Avoid the word we when describing your contribution.

What are the most common mistakes candidates make in this round?

Five repeat mistakes from candidate reports. One, reciting textbook OOP definitions without a code example or a project anchor. Two, writing DSA code from memory but failing to dry-run it on paper with a sample input. Three, claiming React or Spring on the resume but stumbling on useState or dependency injection. Four, naming system design components for the URL shortener but not walking through the redirect happy path. Five, giving a generic why-Digital answer like good company or family pressure instead of a specific reason rooted in client breadth or BFSI scope.

How is this AI interviewer different from a real TCS Digital panel?

Aarav Kumar, the AI persona, models a real Industry Senior Member from the TCS Chennai BFSI Digital practice. He asks one question at a time, waits for a full response, and pushes back twice on vague answers exactly like a campus interviewer. He never praises, never coaches, and never breaks character. The main difference is the rubric is visible to you as a live methodology tracker on the right side of the screen, and you get a transcript-backed scorecard at the end naming the trade-off you could not defend.

How is the scoring done in this practice round?

Six domain metrics and seven score dimensions drive the scoring. Domain metrics include project ownership specificity, DSA code dry-run rigor, system design component coverage, advanced OOP application, advanced DBMS depth, and Digital track motivation articulation. Each metric has scoring anchors from zero to one hundred with example answer phrasing for every band. The methodology tracker on screen ticks off the four must-have signals live during the round. The final report flags the exact moment your defence broke down and gives you the verbatim quote.

What should I do in the first two minutes before the round starts?

Pull up the project you will defend and rehearse the one-line description out loud. Identify the specific module you personally built and the technology choice you made for it. Have one alternative technology ready that you considered and rejected, with one sentence of reasoning. Open a paper or a text doc and pre-draw a linked-list cycle detection trace so your hand is warm. Pull up the URL shortener component list in your head. Plan a specific why-Digital answer that mentions BFSI client breadth or the seven-and-a-half LPA scope, not generic ambition.

How do I handle a question on indexes and isolation levels?

Anchor your answer in a project example. For indexes, explain clustered versus non-clustered with one sentence each, then add when an index hurts write performance. Avoid the textbook recital. For isolation levels, walk through the four standard levels in one breath and tag each with the anomaly it prevents. Dirty read, non-repeatable read, phantom read. If the interviewer pushes deeper, name the trade-off between consistency and concurrency at higher isolation levels. Keep examples concrete and short.

What does a strong answer to the URL shortener question sound like?

A strong answer names at least four components. Load balancer in front, web server replicas, URL shortener service holding generation and lookup logic, and a database storing the short to long mapping. Optionally a cache for hot redirects. Then walk through the happy path. Client sends a long URL, the service generates a short code using base62 or hashing, stores the mapping, and returns the short URL. On redirect, the service looks up the short code, fetches the long URL, returns HTTP 302. No need for scale numbers at fresher altitude.

How is TCS Digital different from Infosys Specialist Programmer for freshers in India?

TCS Digital offers seven-and-a-half LPA, Infosys Specialist Programmer offers nine LPA. The selection gates differ. TCS uses the Advanced Coding section of the National Qualifier Test, Infosys uses the HackWithInfy three-round coding contest. The interview altitude is similar but TCS leans more on project depth and computer science fundamentals while Infosys leans harder on competitive-programming-style DSA. Why-Digital answers should anchor in TCS client breadth, the BFSI Digital practice, and the long learning curve, not in salary comparison.

What if I cannot solve the DSA problem fully in the round?

Pseudocode counts at fresher altitude. Talk through your approach out loud before writing a single line. State the data structure choice and time complexity upfront. If you get stuck, dry-run your partial code on a sample input of three to five elements on paper. The interviewer rewards thought process and edge-case awareness over a perfect compiling solution. Candidates who freeze and refuse to talk lose. Candidates who think aloud while pointing at their paper trace recover even with partial code.

Sources this interview is built on

Real candidate-report URLs (Glassdoor / AmbitionBox / PrepInsta / GeeksforGeeks / Medium) reviewed when authoring the questions, persona, and rubric. Verify the realism yourself.