🎯 Interview & Career · Chapter 5

The behavioural round: STAR stories from your projects

“Tell me about a time you disagreed with someone.” The interviewer doesn't want your opinion on disagreement. They want one real story, in two minutes, where they can see exactly what you did.

Study this lesson free in the app →

The basics

The behavioural round asks about your past behaviour — “tell me about a time when…” — because how you handled real situations is the best predictor of how you'll handle the next ones.

Why it trips up engineers: we answer in generalities (“I always communicate clearly”) or in team terms (“we built a pipeline”). Neither gives the interviewer evidence about you.

The tool — STAR:

• Situation — where, when, what was at stake (one or two sentences)
• Task — what you were responsible for
• Action — what you did, step by step, and why — most of the answer
• Result — what changed, with a number if possible, plus what you learned

The real preparation isn't memorising answers. It's building a story bank: 6–8 true stories from your projects that you can tell well, each able to answer several kinds of question.

Words to know: STAR, story bank, ownership, ‘I’ vs ‘we’, reflection.

Explained

“Tell me about a time you made a mistake.”

Answer A: “I'm a perfectionist, so I sometimes spend too long on details…” — a dodge. Interviewers hear it weekly.

Answer B: “In March I deployed a change to our ride-payouts loader that appended instead of merged. A retry doubled 6,200 rows and drivers were overpaid ₹3.1 lakh. Here's what I did in the next six hours, and the two things I changed afterwards…”

B admits more, and scores far higher. Specific, owned, fixed, learned.

The match report

A sports journalist covering a football match in Shillong doesn't write “the team played well and showed great spirit”.

She writes: “Down 1–0 at half-time, Lyngdoh switched to the left wing, beat two defenders in the 71st minute and crossed for the equaliser.”

The context is one line. The action is specific: who, what, when. The result is a score. You can see the player's decision.

A behavioural answer is a match report about yourself. Vague praise tells the reader nothing; a specific move and a scoreline tells them everything.

Where the two minutes go

0:00        0:25                                    1:35       2:00
│ S + T     │ A C T I O N                            │ RESULT   │
│ context,  │ what I did, in order, and WHY each      │ number + │
│ stakes,   │ step — decisions, trade-offs,           │ learning │
│ my role   │ who I spoke to, what I wrote            │          │
└───────────┴─────────────────────────────────────────┴──────────┘
   ~20%                      ~60%                         ~20%

The most common failure: 90 seconds of Situation (the architecture, the team history), 20 seconds of Action, no Result. Interviewers score the Action.

A full STAR answer, spoken

Q: “Tell me about a time you disagreed with a stakeholder.”

S: “At Meghalaya Cabs the finance team wanted daily revenue to include rides that were completed but not yet paid.” T: “I owned the revenue model, and I thought that would overstate revenue by about 2% on days with payment outages.” A: “Rather than argue definitions, I pulled three months of data and showed the gap per day — it peaked at ₹1.4 lakh during a gateway outage. I proposed two metrics: ‘booked revenue’ including unpaid, and ‘collected revenue’ matching the bank. I wrote both definitions in our metrics doc, and asked the finance lead which one the CFO email should lead with.” R: “They chose collected for the email and booked for ops. Month-end reconciliation queries from finance dropped from about ten to two. What I learned: bring data to a definitions argument, and offer two named metrics instead of one fight.”

≈ 1 minute 50 seconds.

Build a story bank, not answers

List 6–8 real situations from work, internships or serious projects. For each, note: what happened, your actions, the number, the lesson. Then map them:

story                          conflict failure owner ambig deadline influence
────────────────────────────────────────────────────────────────────────────
Duplicate payouts incident        ·       ✓      ✓     ·      ✓        ·
Revenue definition with finance   ✓       ·      ·     ✓      ·        ✓
Migrating cron jobs to Airflow    ·       ·      ✓     ✓      ·        ✓
Late partner data before audit    ·       ·      ✓     ·      ✓        ·
Pushing back on a 1-day deadline  ✓       ·      ·     ·      ✓        ·
Onboarding two junior engineers   ·       ·      ✓     ·      ·        ✓

Every column needs at least one tick. Freshers: final-year projects, internships, hackathons and college teams all count — the structure matters more than the scale.

What makes interviewers write ‘no’

• ‘We’ throughout. They can't tell what you did. Say ‘we’ for the team context, ‘I’ for your actions.
• Hypotheticals. ‘I would talk to them…’ answers a different question. They asked for a time you did.
• Blaming. ‘The analyst was incompetent’ is a red flag even if true. Describe behaviour, not character.
• The fake failure. ‘I work too hard.’ Choose a real mistake with real consequences and real change.
• No result. The story just stops. Always land it: what changed, how much.
• Five-minute epics. Stop at two; they'll ask follow-ups for detail they want.
• Memorised scripts. Sound recited and collapse on the first probe. Know the bullets, not the sentences.

Short setup, long action, a number at the end — and say ‘I’.

Six to eight true stories, each re-angled to answer many questions.

Handling the follow-up probes

After your story, expect probes. Have the answers ready:

• “What would you do differently?” → one concrete change, not ‘nothing’.
• “What did the other person think?” → show you understood their view: “Finance needed one number for the CFO; I'd made it harder.”
• “How did you measure that?” → know where your number came from.
• “What was the hardest part?” → usually a person or a judgement call, not a tool.

If you don't have a story for a question: “I haven't faced exactly that, but the closest was…” — then tell a real adjacent story. Never invent one; probes expose it fast.

Words to know

In two minutes

• Behavioural questions ask for past behaviour because it predicts future behaviour; answer with one specific story.
• STAR: brief Situation and Task, a detailed Action, a concrete Result — about 2 minutes total.
• Action should be roughly 60% of the answer; say ‘I’ and explain why you chose each step.
• Quantify the result where possible — rupees, hours, rows, incidents — and add one line of learning.
• Build a bank of 6–8 real stories covering conflict, failure, ownership, ambiguity, deadline pressure and influence.
• One strong story can answer several questions; practise re-angling it rather than writing new ones.
• Red flags: blaming others, hypotheticals (‘I would…’), no personal action, no result, five-minute rambles.

Test yourself

There are 8 questions, 3 puzzles and 5 recall cards for this lesson in the app — the recall cards come back on later days, just before you would forget.