Data Scientist 1(Instamart) at Swiggy

How it went

What they would tell you

How to prepare

Round by round

1. 1

   Coding60 min

   This round tested my programming fluency and problem-solving ability, with a practical slant. It included:

   • Python Concepts: Functions, error handling (try-except), decorators, multithreading
   • Pandas: Operations like groupby, filtering, and aggregations
   • SQL: Joins, subqueries, filtering conditions I was asked the "Best Time to Buy and Sell Stock" problem. I was given an array of stock prices and had to find the maximum profit from a single buy-sell transaction. I first explained the brute-force approach, then optimized it using a single-pass solution by tracking the minimum price and calculating the max profit at each step. I explained my thought process clearly, and the interviewer appreciated my problem-solving and communication skills.
   • Best Time to Buy and Sell Stock
2. 2

   ML Depth60 min

   It went deep into both theory and practical understanding. Statistics & Probability:

   • Mean, variance

   • Central Limit Theorem

   • Type I vs Type II errors

   • Why sample std dev = σ/√n

   • Proof of variance of sample mean

   • Probability-based what-if questions ML Theory:

   • Precision, Recall, F1, ROC-AUC

   • Regularization: L1 vs L2

   • Bias-variance tradeoff

   • Metrics for unsupervised models (KMeans) Deep Learning:

   • CNN output dimension formula

   • Dropout behavior (stacked vs between layers)

   • Vanishing gradients in RNNs & how LSTM mitigates it

3. 3

   Problem Solving / ML System Design60 min

   This round focused on real-world applications and system-level thinking in data science. We started with a discussion on my past projects, followed by two practical challenges 1.Dynamic Filter Generation:

   • Given a search input, how would you dynamically display relevant filters on the product page?

   • Discussed feature engineering, historical trends, context-aware personalization 2.ETA Prediction Model:

   • Inspired by Swiggy’s delivery time prediction blog

   • What features to consider, how to approach feature drift, fallback logic, real-time vs batch models This round was open-ended, and they were looking for product-focused thinking along with applied ML rigor.

4. 4

   Hiring Manager60 min

   This round was more of a casual but meaningful conversation than a typical interview. The focus was on understanding my mindset and how I’d fit into the team culture. We talked about how I make decisions in unclear situations, my experience with team dynamics, and how I handle feedback. There was also a discussion around how well I align with Swiggy’s values and why I’m specifically interested in joining the Instamart team.

What came up