Jobra
🐞Step 1 of Jobra:AI Debugging Rounds
DSA+AIPreparing you for both.
JOBRA · AI DEBUG ROUND (PREVIEW)sum.cpp
// buggy snippet
int total(vector<int>& a, int n) {
    int sum = 0;
    for (int i = 0; i <= n; i++) { // 🐞        sum += a[i];
    }
    return sum;
}
A real codebase. A real bug. You + AI.

Jobra drops you into a broken repo. Find it, fix it, ship it. AI helps you debug like on the job.

Prompt:Correct the loop condition
Tap 'Let Jobra fix it' below to simulate AI debugging...

Amazon Data Scientist Interview Guide: Coding, ML Fundamentals and Leadership Principles

3 min read

Prepare for Amazon data scientist interviews with reported assessment patterns, interview questions, and preparation advice.

In this article
On this page

Last reviewed November 15, 2026.

For data scientist interns, new grads and experienced candidates.

Built by the Codinzhub team from Amazon data science interview experiences that candidates shared, including the ones posted on Codinzhub. Each experience is one person's interview, so use this as a map, not a promise.

The 30-second version. A screen, then two to five rounds. Most rounds mix Leadership Principle questions with coding (pandas or timed coding-problem), statistics and ML fundamentals, and a deep project discussion. New grads had a hiring manager round with an ML design question. Interns with economics backgrounds could choose a causal inference track.

Round-by-round breakdown

Stage 1: Application, where candidates applied

How it works. Candidates applied online, including PhD students in econometrics. Campus candidates in India interviewed for data science intern roles in 2026.

Stage 2: Screen

At a glance

  • Length: about an hour
  • Questions per round: 4 LP questions and 10 or more ML basics questions
  • What was asked: Bias for Action, Ownership, learning a new skill; gradient descent, regularisation, boosting, transformers
  • Most common: ML fundamentals from scratch

Questions candidates reported:

Question (our summary)What it tests
Gradient descent vs SGD; batch size; why SGDOptimisation
Linear and logistic regression from scratch; the logistic lossModels
Overfitting, bootstrapping and boostingML basics
How a transformer differs from an RNN; what an embedding isDeep learning

More questions: Browse Amazon questions.

Stage 3: The interview rounds

Coding round

At a glance

  • Length: about an hour
  • Questions per round: 1 or 2 problems
  • What was asked: pandas, stacks, graphs, SQL-style questions
  • Most common: Daily Temperatures; number of islands

Questions candidates reported:

Question (our summary)What it tests
Days until a warmer temperatureMonotonic stack
Number of islands; two string and hash map questionsGraphs, hashing
Top 5 users by spending in six months; top product per categorySQL or pandas
Pandas grouping, aggregation and mergingPandas

More questions: Browse Amazon questions.

ML and project rounds

At a glance

  • Length: about an hour
  • Questions per round: a project deep dive and 6 to 10 concept questions
  • What was asked: RAG design, bias and variance, regularisation, dropout, metrics, A/B tests
  • Most common: precision vs recall; your most recent project

Questions candidates reported:

  • Explain a confusion matrix to high-school students; precision, recall, F1, AUC; multi-class evaluation.
  • Bias-variance trade-off; regularisation and dropout in depth, including inference.
  • Batch normalisation; Isolation Forest vs Random Forest; t-tests and A/B tests.
  • How a RAG system works, how to fix bad retrieval, and redesign it when requirements change.
  • Spam detection as an ML design (hiring manager, new grad).

LP round

Questions candidates reported: a calculated risk; extending a deadline; a fast decision; convincing teammates who disagree; leading a team; learning something new; why Amazon.

More questions: Browse Amazon questions.

The Codinzhub Next-Round Playbook

What data scientists who moved on did. Put together by the Codinzhub team from the DS experiences in this guide.

  • Revise ML fundamentals until you can explain them simply, starting from regression.
  • Practise pandas grouping and SQL-style ranking questions.
  • Prepare one project you can redesign live when requirements change.
  • Treat LP questions as the most important part; candidates said so.

About this guide

This guide is researched and written by the Codinzhub team. We read every Amazon data science interview experience we could find, kept only what candidates actually described, and rewrote it in our own words. Found something out of date? Share your own experience on Codinzhub and we will update the guide.

Ready for the next topic?

Keep track of what you’ve learned.