Curated interview practice questions with answers. These are practice scenarios, not a verified list of past interview questions.
About Amazon hiring for Data Science roles
Amazon is a global technology company spanning e-commerce, cloud computing through AWS, devices, advertising and streaming. In India it runs large engineering, operations and analytics teams supporting both Amazon.in and global businesses. It hires for software development engineers, SDETs, quality assurance engineers, business intelligence engineers, data engineers and data scientists.
For Data Science roles, this Amazon practice set focuses most on A/B Testing & Experimentation (8), MLOps and Model Deployment (7), Machine Learning (7), Python for Data Science (6). It has 2 easy, 25 medium, 13 hard questions, so you can start easy and work up to the harder ones.
Company type
Product
Headquarters
Seattle, USA
Main India offices
Bengaluru, Hyderabad, Chennai, Pune, Gurugram
Amazon Data Science interview process
Online assessment (coding + work style survey)
Technical round 1 (DSA)
Technical round 2 (DSA or system design)
Bar raiser round
Hiring manager round
Typical pattern based on public candidate experiences. Rounds vary by role, team and year.
Solved for AmazonStart practicing
3 QUESTIONS FREESign up to unlock the full question bank and track your progress.Unlock every question and answer, then continue practising company-wise and package-wise.
As the number of independent trials grows, the sample average converges to the true expected value, which is why larger samples give more stable estimates.
02
A Pandas merge suddenly returns many more rows than the left table had. What happened?
The join key is probably not unique on the right side, so each left row matches several right rows and the result multiplies. Check with right['key'].duplicated().sum(). Other causes are null or blank keys matching each other after cleaning, or joining on the wrong column. Fix it by deduplicating or aggregating the right table first, and use merge(..., validate='many_to_one') so Pandas raises an error when the assumption breaks.
Use the provider's structured output or JSON schema mode when available, which constrains the output to your schema. Otherwise, describe the exact schema in the prompt with an example, and use low temperature. Always parse and validate the response in code, for example with Pydantic, and on failure retry with the error message included. Keep schemas small and flat, and use enums for fields with fixed values.
04
What are shadow, canary and A/B deployments for models?
In shadow deployment the new model receives real traffic and logs predictions, but users still get the old model's output; it is safe for checking latency and agreement. A canary sends a small share of traffic, like 5 percent, to the new model and expands if metrics stay healthy. An A/B test splits traffic randomly to measure business impact statistically. Teams often go shadow, then canary, then full rollout.
05
Your power analysis says you need 200,000 users per group, but the page gets only 5,000 users a week. What are your options?
Waiting 80 weeks is unrealistic. Options: test a bolder change so the expected effect is larger; use a more sensitive metric closer to the change, like add-to-cart instead of purchase; apply CUPED to reduce variance; or move the test to a higher-traffic page. If none work, accept a larger MDE explicitly, or decide using qualitative research and a staged rollout with monitoring instead of a formal test.
06
Write Python code to group a list of order dictionaries by city and sum the amounts without Pandas.
RFM stands for recency, frequency and monetary value. From the transactions table, for each customer as of a cut-off date: recency is days since the last purchase, frequency is the number of orders in the last 90 days, and monetary is total or average spend in that window. Compute them using only data before the cut-off, add trends like spend this month versus last, and the label from the period after.
Run an A/B/n test with control plus five variants and equal random allocation. Calculate sample size for the smallest effect worth detecting, remembering that each variant gets only a sixth of traffic, so the test takes longer. Plan the analysis upfront: compare each variant to control and adjust for multiple comparisons. If the team has strong priors, test fewer designs first to save time.
10
What is CUPED and how does it reduce variance in experiments?
CUPED uses a pre-experiment covariate, usually the same metric for each user in the weeks before the test, to remove predictable variation. The adjusted metric is Y minus theta times (X minus the mean of X), where theta is cov(X, Y) divided by var(X). Because randomisation makes X balanced across groups, the estimate stays unbiased while variance can drop by 30 to 50 percent, letting tests reach significance faster with the same traffic.
11
How does a multi-armed bandit differ from a classic A/B test?
A classic A/B test splits traffic in fixed proportions until the end, then decides, which is best for learning and clean inference. A multi-armed bandit, such as Thompson sampling, shifts traffic towards better-performing variants as data arrives, reducing the cost of showing a weak variant. Bandits suit short-lived decisions like headlines or promotions. They make unbiased effect estimates and guardrail analysis harder, so they are poor for major product decisions.
12
Write pytest unit tests for a feature function that computes days since last purchase.
WITH steps AS (
SELECT user_id,
MAX(CASE WHEN event = 'visit' THEN 1 ELSE 0 END) AS visited,
MAX(CASE WHEN event = 'add_to_cart' THEN 1 ELSE 0 END) AS carted,
MAX(CASE WHEN event = 'purchase' THEN 1 ELSE 0 END) AS purchased
FROM events
WHERE event_date >= '2026-10-01'
GROUP BY user_id
)
SELECT SUM(visited) AS visits, SUM(carted) AS carts, SUM(purchased) AS buyers,
ROUND(100.0 * SUM(carted) / NULLIF(SUM(visited), 0), 1) AS visit_to_cart_pct,
ROUND(100.0 * SUM(purchased) / NULLIF(SUM(carted), 0), 1) AS cart_to_buy_pct
FROM steps;
14
How do you evaluate and explain a model to business stakeholders?
Tie metrics to business outcomes such as cost saved, show feature importance or SHAP values, give example predictions and be transparent about limitations.
15
What are the novelty effect and the primacy effect in A/B tests?
The novelty effect is when users engage more with a change simply because it is new, so early results overstate the long-term lift. The primacy effect is the opposite: existing users resist change at first, so early results understate it. Both fade with time. To detect them, plot the treatment effect by day and by user tenure, compare new users with existing users, and run the test long enough for behaviour to settle.
16
How do you work with domain experts to find good features?
Situation: on an insurance claims model, my first features gave average results. Task: improve fraud detection. Action: I sat with two claim investigators for a day and asked what made them suspicious. They mentioned claims filed soon after policy start and repeated garage names. I turned these into features and checked them with the team. Result: recall at the same precision rose by 12 points. Experts often know signals the data hides.
17
You need to extract fields from scanned invoices that come in hundreds of different layouts. How would you design it?
Fixed templates will not scale to hundreds of layouts. Start with preprocessing: deskew, denoise and detect text regions. Run OCR to get words with their positions. Then use a layout-aware document model, or a vision-language model, to map words to fields like invoice number, GSTIN and total. Add validation rules, such as GSTIN format and line items summing to the total, and send low-confidence documents to human review.
18
Your image classifier trained on 2,000 images memorises the training set within 5 epochs. What do you try?
With so little data, I would use transfer learning from a pretrained model like ResNet or EfficientNet and fine-tune only the top layers first. I would add data augmentation such as flips, crops and colour changes, use dropout and weight decay, and apply early stopping on validation loss. Collecting more labelled images or using a smaller model also helps. I would check for duplicate images across train and validation sets too.
19
The business asks you to pick a decision threshold for a churn model. How do you choose it?
The default 0.5 is rarely right. I would get the cost of each outcome: the value of saving a churner, the cost of a retention offer, and the team's capacity to contact customers. Then I compute expected profit or precision and recall at many thresholds on validation data and pick the one that maximises value within capacity. I would also check that the chosen threshold stays stable over recent months.
20
What are embeddings and vector databases, and why do LLM apps use them?
An embedding model turns text into a vector so that similar meanings are close in space. A vector database stores these vectors with metadata and finds the nearest ones to a query vector quickly using approximate nearest neighbour indexes like HNSW. LLM apps use them for semantic search and retrieval-augmented generation, finding the most relevant documents to put into the prompt. Examples include pgvector, FAISS and dedicated vector stores.
Start with a lightweight one-stage detector at the smallest input size that keeps acceptable accuracy. Optimise it: export to ONNX, then use a hardware runtime like TensorRT or OpenVINO, with FP16 or INT8 quantisation calibrated on real images. Prune or distil from a larger model if needed. Profile the full pipeline, including decoding and preprocessing, and run detection every few frames with tracking in between if compute is tight.
23
Tell me about a time you stopped a team from shipping a change based on a flawed experiment.
Situation: a team wanted to launch a feature after a test showed a 6 percent lift. Task: as the reviewing data scientist, validate it. Action: I found that assignment was triggered after users clicked into the feature, so treatment had only engaged users. I explained the selection bias simply, proposed assigning at app open, and helped rerun it. Result: the true lift was under 1 percent, and we added assignment checks to the review template.
It projects data onto orthogonal directions of maximum variance to reduce dimensions, remove correlation and speed up models, at some cost of interpretability.
Gradient boosting builds trees one after another. It starts with a simple prediction, like the mean. Each new small tree is fitted to the negative gradient of the loss, which for squared error is just the residuals. Its predictions are multiplied by a learning rate and added to the model. Repeating this slowly reduces error. Key settings are number of trees, learning rate, tree depth and subsampling, which together control overfitting.
26
What are precision and recall trade-offs in practice?
Raising the threshold improves precision but lowers recall. Fraud detection leans to recall, while spam filters often prefer precision; choose by the cost of each error.
A forest averages many trees trained on bootstrapped rows and random feature subsets, which cuts variance and overfitting at the cost of interpretability.
Keep notebooks for exploration and reporting only. Move reusable logic into a Python package with modules for data loading, features, training and evaluation, and import it into notebooks. Keep settings in config files, write unit tests for feature functions, and use pre-commit with a formatter and linter. Track experiments in a tool like MLflow. A clear folder template and code reviews keep the team consistent as the project grows.
30
Your fraud model's true labels arrive 30 days late. How do you monitor it in production?
Since accuracy cannot be measured quickly, monitor proxies. Track input data drift with PSI per feature, prediction score distribution, the share of transactions flagged, and data quality like missing values. Watch early signals such as chargeback complaints or manual review outcomes on a sample. Set alerts on sudden shifts. When labels arrive, compute real precision and recall by cohort and compare them with the proxy alerts to tune them.
31
You want to test a new pricing algorithm in a two-sided marketplace where buyers and sellers interact. Why is user-level randomisation a problem and what would you do?
Treated and control buyers compete for the same sellers' supply, so a treatment that makes treated buyers buy more takes inventory from control buyers, exaggerating the measured lift. Use designs that limit interference: switchback tests that alternate the whole market between variants over time slots, or randomise by city or region. Analyse at the switch or cluster level with appropriate variance, and accept that these tests need longer to reach power.
32
Your RAG chatbot often answers using the wrong documents. How do you fix it?
First check retrieval on its own with a labelled set of questions, since generation cannot fix bad context. Then improve chunking so passages are self-contained, add metadata filters such as product or date, try a better embedding model, and add hybrid search with keywords. A reranker on the top 20 results often helps most. Also rewrite vague user questions into clearer search queries before retrieval.
33
How do you choose a learning rate for a neural network?
Chunking splits documents into smaller pieces before embedding them, so retrieval returns focused passages that fit in the prompt. Small chunks, like 200 tokens, are precise but lose context; large chunks, like 1,000 tokens, keep context but add noise and cost. Split on natural boundaries such as headings and paragraphs, add some overlap, and keep metadata like title and section. Then tune chunk size using retrieval metrics on real questions.
35
Your sales forecast scores well in k-fold cross-validation but fails in production. What could be wrong with the validation?
Random k-fold mixes future and past rows, so the model trains on data from after the period it predicts. That leaks information and inflates scores. Use time-based splits instead: train on an expanding or rolling window and validate on the next period, like TimeSeriesSplit in scikit-learn. Leave a gap between train and validation when features use lagged values, so the test reflects real forecasting conditions.
36
Halfway through a test you discover a bug that broke the treatment for Android users. What do you do?
Stop or pause the affected part, fix the bug and restart the test with fresh data, since mixing broken and fixed periods biases results. Do not simply drop Android users from analysis afterwards, because that can break randomisation balance if the bug affected behaviour or logging. Document the incident, and add pre-launch checks such as running the variant through QA on every platform and monitoring guardrails in the first hours.
37
What is the difference between generative and discriminative models?
Discriminative models learn P(y given x) directly (logistic regression, SVM). Generative models learn the joint P(x, y) and can generate data (Naive Bayes, GANs).
38
What is the difference between saving a model with pickle, joblib and ONNX?
pickle serialises Python objects; it is simple but tied to Python and library versions, and loading untrusted pickle files is a security risk. joblib is similar but more efficient for objects with large NumPy arrays, like scikit-learn models. ONNX is an open, language-neutral format that runs in optimised runtimes across languages and hardware, useful for fast serving. Always store library versions alongside pickle or joblib files.
39
How do you use groupby with multiple aggregations?
from pydantic import BaseModel, Field, ValidationError
class LoanRequest(BaseModel):
age: int = Field(ge=18, le=100)
income: float = Field(gt=0)
city: str = Field(min_length=2)
def predict(payload: dict):
try:
req = LoanRequest(**payload)
except ValidationError as e:
return {'error': e.errors()}
return {'score': float(model.predict_proba([[req.age, req.income]])[0, 1])}
Tips to crack the Amazon Data Science interview
Prepare STAR-format stories for each Amazon Leadership Principle, such as Customer Obsession, Ownership and Dive Deep.
Practise medium-level DSA problems on arrays, trees, graphs and dynamic programming, explaining time and space complexity aloud.
For BIE or data roles, practise SQL window functions and metric-driven case questions on e-commerce funnels and delivery operations.
Keep a one-minute story ready for every project on your CV: problem, your part, tools, result.
Practise the hard questions on this page aloud; Amazon interviewers usually ask follow-ups on your first answer.
Frequently asked questions
How many rounds are there in the Amazon Data Science interview?
Candidates usually go through 5 stages: Online assessment (coding + work style survey); Technical round 1 (DSA); Technical round 2 (DSA or system design); Bar raiser round; Hiring manager round. The exact rounds change with the role, team, location and your experience, so confirm the process with the recruiter.
Which topics should I prepare for the Amazon Data Science interview?
This practice set leans on A/B Testing & Experimentation (8), MLOps and Model Deployment (7), Machine Learning (7), Python for Data Science (6). Revise those first, then practise explaining one or two of your own projects end to end, because most Amazon interviewers ask you to walk through real work.
Are these the exact questions asked at Amazon?
No. These are curated practice questions that match the skills and difficulty usually tested in product interviews like Amazon's. They are not a leaked or verified list of past questions.
Where does Amazon hire Data Science professionals in India?
Amazon has large teams in Bengaluru, Hyderabad, Chennai, Pune, Gurugram. Openings and locations change often, so check the Amazon careers page and filter by city and experience.