Genpact Interview Guide (2026)

1Snapshot, read this first#

What to actually understand about Genpact. It is a process and services company at heart, and that shapes every round.

• Its roots are in GE's back-office operations, which it spun out of in 2005. Lean and Six Sigma still run through everything.
• It does not need you to out-code a product-company candidate. It needs someone who thinks clearly, communicates cleanly, and connects technical work to a business outcome.
• Every round is really testing one question: can we put this person in front of a client's process next month and trust them not to embarrass us?

Every answer you give is auditioning for a seat in front of a client's process. Talk like it.

2Getting the call (the part most guides skip)#

Volume hiring runs through a few channels, and each has its own door. Pick the one that fits you and work it hard instead of spraying applications everywhere.

Campus and off-campus drives. Most fresher hiring flows through these.

• They are advertised on the Genpact careers site and on portals like Naukri and AmbitionBox listings.
• When a drive opens, register in the first 24 to 48 hours. These windows fill and close fast.
• Use a clean personal email (firstname.lastname@gmail.com), not a college ID that may expire.
• Fill the form with no typos in your name, phone, or degree. Miss a field and you may never get the test link.
• The "application" and the "first round" are often the same event: these feed straight into the online aptitude test.

Referrals. A referral moves your resume past the first automated filter and signals reliability, which matters a lot where attrition is the enemy.

• Ask an insider to submit you through the internal referral portal, not a WhatsApp forward.
• A portal referral gets an employee ID attached to it. A forwarded PDF gets lost.
• Give them a 3-line blurb they can paste, so it is zero effort for them.

Portals and recruiters. For lateral roles, recruiters and staffing partners search by keyword.

• Keep LinkedIn and Naukri current with the exact skill names the JD uses: SQL, Python, PySpark, Power BI, Tableau, AWS, Azure.
• "Data visualization" and "Power BI" are not interchangeable to a search filter. List both the category and the specific tool.

Resume rules. Keep it one page for freshers, two max for laterals.

• Lead every bullet with a verb and a number.
• Weak: "Worked on a dashboard for the sales team."
• Strong: "Built a Power BI dashboard pulling from 3 SQL sources that cut weekly sales-reporting effort from 6 hours to 30 minutes for a 12-person team."
• That one bullet shows a tool, a data reality, and a measurable business win, which is exactly the frame Genpact rewards.
• If a bullet has no number, find the number or cut the bullet.

ATS keywords. Mirror the job description.

• If it says "SQL, data visualization, ETL, stakeholder communication," those exact words should appear naturally, not paraphrases.
• Spell out skills plainly in a dedicated skills line. The parser and the recruiter both scan in seconds.
• Save and send as a PDF named Firstname_Lastname_Resume.pdf so it survives the upload intact.

3The loop at a glance#

Not everyone sees all five as separate sittings.

• Technical and case are sometimes merged into one 45-minute conversation.
• For high-volume fresher drives, the aptitude test plus a short communication check is often the main gate before a single combined human round.

Here is the typical flow:

4The technical rounds, patterns, not problems#

Genpact does not run a LeetCode-hard gauntlet. The bar is "can you reason correctly and explain it," not "can you solve a hard dynamic-programming problem under a 20-minute timer." Prepare for patterns, not a 300-problem bank.

The aptitude test is the real filter. This is where the majority of the funnel gets cut, so treat it seriously.

• Quantitative: percentages, ratios and proportion, averages, time-speed-distance, time and work, profit and loss, simple and compound interest, permutations/combinations basics, number systems. Standard placement-aptitude topics, the same ones you would drill for TCS NQT or an Infosys drive.
• Logical reasoning: seating arrangement, blood relations, syllogisms, series completion, coding-decoding, direction sense, puzzles.
• Verbal / English: reading comprehension, sentence correction, synonyms/antonyms, error spotting, para-jumbles.
• Data interpretation: reading tables, bar and pie charts, and doing quick arithmetic on them without a calculator.

Practice timing above everything.

• The questions are not individually hard. The pressure is doing many of them accurately and fast, often around a minute per question.
• The common failure is not wrong answers, it is unanswered ones.
• Build the habit: skip a stuck question in 45 seconds, flag it, come back.
• Aim to leave no easy mark on the table.

The technical/pseudocode section (tech and analytics roles). What you see depends on the role.

• SQL: joins, GROUP BY with aggregates, HAVING, subqueries, window functions (ROW_NUMBER, RANK, DENSE_RANK), finding the Nth-highest value, removing duplicates.
• Python / programming logic: string and list manipulation, loops, dictionaries, basic file handling, reading a problem and writing clean pseudocode. For analytics, pandas basics (filter, groupby, merge) help. They care more that your logic is correct and readable than that you used the slickest one-liner.
• CS fundamentals (covered in Section 5) for software roles.
• Analytics/stats basics: mean/median/mode, standard deviation, correlation vs causation, a plain-English read on what a dataset is telling you. Expect "when would you use median over mean?" and a quick sanity check on an outlier.

How to win the round. The content is only half the score. The other half is how you work a problem, and services interviewers grade that explicitly.

1. Clarify first. Restate the question in your own words and confirm assumptions before writing anything. "So the table has one row per transaction, and I need total revenue per customer for last quarter, correct?" This alone separates you from the candidate who writes the wrong query fast.
2. Think out loud. Narrate your approach step by step. A silent correct answer scores lower than a spoken one here, because the client-facing instinct is what they are buying.
3. State complexity and tradeoffs. Even a simple "this is a single pass, so linear time" or "a subquery works but a window function is cleaner and reads better" shows maturity.
4. Test your answer. Walk one example row through your SQL or logic out loud. Catching your own edge case (a NULL, a tie, an empty group) before the interviewer points it out is one of the strongest signals you can send.

5CS fundamentals and the case/scenario round#

For software roles, Genpact leans on CS fundamentals far more than heavy system design. Have crisp, example-backed answers ready, a 10-second plain-English version and one concrete example for each.

• DBMS: normalization (1NF to 3NF), primary vs foreign keys, indexes and why they speed reads but slow writes, ACID properties, the join types, SQL vs NoSQL and when each fits. Be ready to write a query live, not just define terms.
• OS: process vs thread, deadlock and its four conditions, scheduling basics (FCFS, round robin), paging vs segmentation at a high level.
• Networking: OSI layers in brief, TCP vs UDP with a real example of each, what actually happens when you type a URL and hit enter (DNS, TCP handshake, request, response, render), common HTTP status codes like 200, 404, 500.
• OOP: the four pillars (encapsulation, abstraction, inheritance, polymorphism) each with a one-line real example, plus overloading vs overriding and why the difference matters.

Do not memorize definitions you cannot use.

• "ACID means atomicity, consistency, isolation, durability" scores nothing on its own.
• "ACID is why a bank transfer either fully happens or fully rolls back, so you never debit one account without crediting the other" lands, because it shows you know what the concept is for.

The case / scenario round. This is where Genpact's services DNA shows.

• You may get a loosely defined situation: "A client's monthly report is taking too long to produce and has errors. How would you investigate and fix it?"
• They are not looking for the one right answer. They want structure, and they are watching whether you jump to a solution or think first.

Work it in four moves:

1. Clarify scope: what report, how long is "too long," which errors, who is affected, how often it runs.
2. Break the problem into buckets: data sources, the process steps, the people doing them, the tooling. Keep the buckets non-overlapping so you look organized.
3. Form a hypothesis: "errors likely come from a manual copy-paste step; slowness likely from an un-indexed query or a full-table scan on a growing dataset."
4. Propose how you would validate and fix, and always close on business impact: "I would time each step first to confirm where it is slow, then fix the top one. If my hunch is right, this frees up roughly two analyst-days a month and removes the error class entirely."

Thinking in that MECE-style, impact-aware way is exactly the consulting instinct Genpact is hiring for.

6Behavioral and culture fit#

Genpact's stated purpose is "the relentless pursuit of a world that works better for people," paired with values commonly summarized as curiosity, courage, and incisiveness.

• Curiosity: being genuinely curious about a client's problem.
• Courage: having the nerve to challenge the status quo.
• Incisiveness: being sharp and decisive.
• Plus a continuous-improvement mindset: the Lean and Six Sigma roots from the GE days still run deep.

Treat those three words as the lens you are scored against, and make sure at least one STAR story visibly hits each.

Prepare STAR stories (Situation, Task, Action, Result) for:

• A time you solved a hard problem.
• A conflict with a teammate.
• A failure and what you learned.
• A time you improved a process.
• A time you dealt with an unhappy stakeholder.
• Always end on a number. "It went better" is not a result. "It dropped from 3 days to same-day" is.

One full worked STAR example:

• Situation: During my final-year internship, the team's weekly inventory reconciliation report was built by hand in Excel and regularly shipped 2-3 days late with a handful of mismatched totals.
• Task: I was asked to make the report reliable and on time, without any new budget for tools.
• Action: I mapped the manual steps, found that two analysts were pulling from different source extracts, standardized on one SQL query as the single source of truth, and built a templated workbook with validation checks that flagged mismatches automatically. I documented the steps so anyone on the team could run it without me.
• Result: Report delivery moved from 2-3 days late to same-day, and reconciliation errors dropped to near zero across the next eight weeks, saving roughly 6 analyst-hours a week.

Why that story works.

• Curiosity: dug into the root cause instead of patching symptoms.
• Courage: changed an entrenched manual process two people were attached to.
• Incisiveness: one source of truth, decided fast.
• Continuous improvement, with a measurable result.
• Have two or three stories this tight so you are never scrambling for a number mid-answer.

7Your prep plan#

4-week plan

• Week 1: Lock aptitude fundamentals, one topic set per day (quant, reasoning, verbal), 30 timed questions daily. Track your accuracy and your time per question, not just whether you finished.
• Week 2: Core technical, SQL every day (joins, then aggregates and HAVING, then window functions and Nth-highest), plus Python/pseudocode or CS fundamentals for your role.
• Week 3: Case practice and STAR stories, write 6 STAR stories mapped to curiosity/courage/incisiveness, do 3 mock scenario questions out loud.
• Week 4: Full mock loops, one timed aptitude test, one technical mock, one HR mock. Fix weak spots. Polish spoken English daily by recording yourself and listening back.

2-week plan

• Days 1-4: Timed aptitude every day; SQL fundamentals in parallel.
• Days 5-8: Technical fundamentals (SQL window functions, Python basics or DBMS/OS/OOP) plus one case per day.
• Days 9-12: Write and rehearse 6 STAR stories; one mock HR round.
• Days 13-14: Two full mock sittings; rest and review notes.

2-day crash plan

• Day 1 AM: One timed aptitude test; list your weakest 2 topics and drill only those.
• Day 1 PM: SQL joins, aggregates, and Nth-highest; one case question out loud; draft 4 STAR stories.
• Day 2 AM: Role fundamentals quick-review (SQL or CS basics); rehearse STAR stories aloud until they are tight and timed under two minutes each.
• Day 2 PM: Mock HR Q&A, practice "tell me about yourself" and "why Genpact," confirm your comp and relocation answers so you are not improvising them, sleep early.

8Why people get rejected at Genpact#

1. Weak spoken English. For a client-delivery company this is often an immediate no, regardless of how strong the resume is. Fluency and clarity matter as much as content, and there is a dedicated test for it.
2. Aptitude test fails on speed. Candidates know the topics but run out of time and leave easy marks unclaimed. The fix is timed full tests, not untimed topic practice.
3. Memorized answers that collapse. A rehearsed definition with no real understanding falls apart on the first "okay, but why?" follow-up. Interviewers probe exactly there to find the fakes.
4. No business framing. Technically correct but cannot say why the work mattered to anyone. This is the opposite of the consulting instinct they are hiring for.
5. Vague project ownership. "We built it" with no clear "here is what I specifically did, here is the decision I made, and here is what it achieved." Services interviewers push on the "I" relentlessly.
6. Inflexibility on basics. Being unwilling to even discuss shifts, location, or relocation can quietly end a services-role candidacy, because delivery centers run on that flexibility.

9The offer, comp and negotiation#

The ranges, honestly.

• Freshers: entry roles (Process Associate, analyst, developer) commonly land in the ~₹3.5-6.5 LPA range, with Management Trainee and niche-skill roles higher.
• Laterals: tech/analytics roles span roughly ₹8-20+ LPA depending on band, skill, experience, and city. Cloud, data engineering, and AI/ML command the top of that range.
• Packages typically include a fixed base, a variable/performance component, and standard benefits. Some delivery and night-shift roles carry a shift allowance on top.

Negotiation tips for Genpact's context:

1. Anchor on skill and market, not need. If you bring an in-demand stack (PySpark, a cloud platform, strong SQL plus a visualization tool), name comparable market ranges and ask where in the band you would sit: "Based on what I am seeing for this skill set, I was expecting somewhere around X, where does that land in your band?"
2. Negotiate the whole package. If base is banded and rigid, which it often is in a large services firm, push on joining bonus, band/level, or an earlier first review cycle rather than only on base.
3. Be flexible where it is cheap for you. Willingness on shift or location can unlock a better offer. Trade it deliberately for something you want rather than conceding it for free.

Be polite, specific, and backed by a competing offer or clear market data. Services firms respond to reasoned asks, not ultimatums, and the recruiter you negotiate with may well be the person who helps you later, so keep it warm.

Want this turned into a personalized, dated prep schedule around your actual interview date? Build your own week-by-week plan from the sections above and start with whichever round you are weakest on, usually aptitude speed or spoken English.