Question 1

A colleague says you cannot use a t-test because your revenue metric is heavily right-skewed. How do you respond, and what sample size would make you comfortable proceeding?

Accepted Answer

The Central Limit Theorem guarantees that the sampling distribution of the mean approaches normality as sample size grows, regardless of the underlying distribution. For moderately skewed distributions, n=30 is often cited as a rule of thumb, but for heavy-tailed metrics like revenue, n=200 or more may be needed. The candidate should mention that they would check with a bootstrap distribution of the mean to verify normality, and consider log-transforming or using a non-parametric test like Mann-Whitney U as an alternative.

Question 2

Explain intuitively, without relying on equations, why adding a penalty term to the loss function prevents a model from memorizing the training data.

Accepted Answer

Regularization penalizes model complexity, forcing the model to 'earn' each increase in weight by a corresponding improvement in training loss. Without it, the model can freely grow weights to fit even noise in the training data. With a penalty, the optimizer will only make weights large when the signal in the data is strong enough to outweigh the penalty cost, so the model learns the true pattern rather than memorizing individual examples.

Question 3

Your model shows that income strongly predicts health outcomes, but income is correlated with education. Is education a confounder or a mediator here, and why does the distinction matter for your analysis?

Accepted Answer

Education is a confounder if it is a common cause of both income and health but is not on the causal path from income to health. It is a mediator if income causes higher education which in turn causes better health, so income affects health partly through education. The distinction matters because conditioning on a mediator blocks the very causal path you are trying to estimate, producing what is sometimes called over-adjustment bias. Conditioning on a confounder, by contrast, is necessary to remove bias. Mistaking a mediator for a confounder and controlling for it will underestimate the total effect of income on health. The candidate should reference a directed acyclic graph (DAG) as the right tool for making this determination.

Question 4

You are choosing between Random Forest and XGBoost for a tabular classification problem. Walk me through the factors that would make you choose one over the other.

Accepted Answer

Random Forest trains trees in parallel (faster for large datasets), is more robust to hyperparameter settings, handles missing values less gracefully (requires imputation), and is harder to overfit. XGBoost typically achieves higher accuracy on structured data through sequential error correction, handles sparse data and missing values natively, but is more sensitive to hyperparameter tuning and slower to train. Choose Random Forest when training time or tuning budget is limited; choose XGBoost when maximizing predictive performance with sufficient tuning time.

Question 5

Early neural networks used sigmoid activations but modern deep networks use ReLU and its variants. Explain why this change happened and when you would still use sigmoid.

Accepted Answer

Sigmoid saturates at both ends, causing vanishing gradients in deep networks (gradient approaches zero, weights stop updating). ReLU has a gradient of 1 for positive inputs, solving the vanishing gradient problem for deep networks. However, ReLU has the dying ReLU problem (units stuck at zero for all inputs). Leaky ReLU and ELU fix this. Sigmoid is still used in the output layer for binary classification and in gating mechanisms (LSTM). The candidate should mention that the computational cost of ReLU is also lower.

Question 6

A test for a disease has a 99% sensitivity and 95% specificity. The disease affects 1 in 10,000 people. You test positive. What is the probability you actually have the disease?

Accepted Answer

Using Bayes theorem: P(disease | positive) = P(positive | disease) * P(disease) / P(positive). P(positive) = P(positive | disease) * P(disease) + P(positive | no disease) * P(no disease) = 0.99 * 0.0001 + 0.05 * 0.9999 = 0.000099 + 0.049995 = 0.050094. P(disease | positive) = 0.000099 / 0.050094, which is approximately 0.197%, or roughly 1 in 500. This illustrates base rate neglect. The candidate should walk through this calculation clearly and explain why a highly accurate test can still be mostly wrong when the base rate is very low.

Question 7

In a SQL query that has both a GROUP BY clause and a window function, in what order do they execute, and what does that imply about which values the window function can see?

Accepted Answer

SQL logical execution order: FROM, WHERE, GROUP BY, HAVING, SELECT (window functions evaluated here, after aggregation), ORDER BY. This means a window function in the SELECT clause sees only the grouped rows, not the original pre-aggregated rows. If you want to apply a window function to unaggregated data, it must be done in a subquery or CTE before the GROUP BY. The candidate should illustrate with an example where a running total over monthly grouped rows behaves differently from one over the raw event rows.

Question 8

Why does the curse of dimensionality make k-nearest neighbors unreliable in high-dimensional spaces, and what does this imply about feature engineering?

Accepted Answer

In high dimensions, all points become approximately equidistant from each other because most volume is near the surface of the hypersphere. This means nearest neighbors are no longer meaningfully close, so k-NN predictions degrade. For feature engineering, this implies that adding irrelevant dimensions hurts distance-based models significantly, so dimensionality reduction (PCA, feature selection) is critical before applying k-NN or density-based clustering. It also explains why deep learning models that learn low-dimensional representations outperform naive methods on raw high-dimensional inputs.

Question 9

How would you use expected value to decide whether to ship a feature that showed a 2% lift in revenue but a 1% lift in support ticket volume? Walk through the calculation.

Accepted Answer

Expected value requires quantifying the dollar value of a 2% revenue lift and the cost of a 1% increase in support tickets. The candidate should walk through a concrete example: if annual revenue is 10M, a 2% lift is 200k. If each support ticket costs 15 dollars and volume is 100k tickets per year, a 1% increase is 1,000 tickets or 15k dollars. Net expected value is 200k minus 15k = 185k, ship it. The candidate should also discuss uncertainty around both estimates and the importance of using confidence intervals rather than point estimates.

Question 10

Explain what the attention mechanism in a transformer does in plain terms, without using the words 'query', 'key', and 'value' as jargon. Why is this better than the fixed-length context vector in RNNs?

Accepted Answer

Attention lets each output token look back at all input tokens and decide which ones are most relevant to it, producing a weighted combination of input representations. This is better than RNNs which compress all input information into a single fixed-size vector and then try to decode from it, which loses information for long sequences. The candidate should explain that attention is O(n^2) in sequence length, which is why it became computationally feasible only as hardware scaled, and why sparse attention mechanisms exist.

Question 11

Explain how MapReduce works conceptually, then write the map and reduce functions to count word frequencies across a large corpus of documents that does not fit in memory on a single machine.

Accepted Answer

A strong answer explains that the map phase produces (key, value) pairs from input chunks in parallel across many workers, and the reduce phase aggregates values for each unique key after a shuffle and sort step. For word count, map emits (word, 1) per word and reduce sums the values for each word. The candidate should mention that the shuffle is the bottleneck, explain how combiners reduce network traffic, and note that MapReduce is largely superseded inside Google by FlumeJava (the inspiration for Apache Beam and Cloud Dataflow), but the concept remains foundational and still appears in interview discussions.

Question 12

Snowflake uses columnar storage optimized for analytics. Explain exactly why columnar storage is faster for aggregate queries but worse for point lookups, how data is compressed in columnar format, and when you would still reach for a traditional row store.

Accepted Answer

Columnar storage keeps all values of a single column contiguous on disk, so a SUM(revenue) query reads only the revenue column, skipping all other columns. This dramatically reduces I/O for wide tables. Compression is much more effective per column because values in a column share the same type and often have low cardinality (run-length encoding, dictionary encoding, delta encoding all apply well). For a point lookup by primary key (fetch one row), you must read from each column file to reconstruct the row, making it slower than a row store where one disk seek gets the full row. Row stores excel at high-concurrency OLTP with frequent single-row reads and writes.

Question 13

A colleague says you cannot use a t-test because your revenue metric is heavily right-skewed. How do you respond, and what sample size would make you comfortable proceeding?

Accepted Answer

The Central Limit Theorem guarantees that the sampling distribution of the mean approaches normality as sample size grows, regardless of the underlying distribution. For moderately skewed distributions, n=30 is often cited as a rule of thumb, but for heavy-tailed metrics like revenue, n=200 or more may be needed. The candidate should mention that they would check with a bootstrap distribution of the mean to verify normality, and consider log-transforming or using a non-parametric test like Mann-Whitney U as an alternative.

Question 14

Explain intuitively, without relying on equations, why adding a penalty term to the loss function prevents a model from memorizing the training data.

Accepted Answer

Regularization penalizes model complexity, forcing the model to 'earn' each increase in weight by a corresponding improvement in training loss. Without it, the model can freely grow weights to fit even noise in the training data. With a penalty, the optimizer will only make weights large when the signal in the data is strong enough to outweigh the penalty cost, so the model learns the true pattern rather than memorizing individual examples.

Question 15

Your model shows that income strongly predicts health outcomes, but income is correlated with education. Is education a confounder or a mediator here, and why does the distinction matter for your analysis?

Accepted Answer

Education is a confounder if it is a common cause of both income and health but is not on the causal path from income to health. It is a mediator if income causes higher education which in turn causes better health, so income affects health partly through education. The distinction matters because conditioning on a mediator blocks the very causal path you are trying to estimate, producing what is sometimes called over-adjustment bias. Conditioning on a confounder, by contrast, is necessary to remove bias. Mistaking a mediator for a confounder and controlling for it will underestimate the total effect of income on health. The candidate should reference a directed acyclic graph (DAG) as the right tool for making this determination.

Question 16

You are choosing between Random Forest and XGBoost for a tabular classification problem. Walk me through the factors that would make you choose one over the other.

Accepted Answer

Random Forest trains trees in parallel (faster for large datasets), is more robust to hyperparameter settings, handles missing values less gracefully (requires imputation), and is harder to overfit. XGBoost typically achieves higher accuracy on structured data through sequential error correction, handles sparse data and missing values natively, but is more sensitive to hyperparameter tuning and slower to train. Choose Random Forest when training time or tuning budget is limited; choose XGBoost when maximizing predictive performance with sufficient tuning time.

Question 17

Early neural networks used sigmoid activations but modern deep networks use ReLU and its variants. Explain why this change happened and when you would still use sigmoid.

Accepted Answer

Sigmoid saturates at both ends, causing vanishing gradients in deep networks (gradient approaches zero, weights stop updating). ReLU has a gradient of 1 for positive inputs, solving the vanishing gradient problem for deep networks. However, ReLU has the dying ReLU problem (units stuck at zero for all inputs). Leaky ReLU and ELU fix this. Sigmoid is still used in the output layer for binary classification and in gating mechanisms (LSTM). The candidate should mention that the computational cost of ReLU is also lower.

Question 18

A test for a disease has a 99% sensitivity and 95% specificity. The disease affects 1 in 10,000 people. You test positive. What is the probability you actually have the disease?

Accepted Answer

Using Bayes theorem: P(disease | positive) = P(positive | disease) * P(disease) / P(positive). P(positive) = P(positive | disease) * P(disease) + P(positive | no disease) * P(no disease) = 0.99 * 0.0001 + 0.05 * 0.9999 = 0.000099 + 0.049995 = 0.050094. P(disease | positive) = 0.000099 / 0.050094, which is approximately 0.197%, or roughly 1 in 500. This illustrates base rate neglect. The candidate should walk through this calculation clearly and explain why a highly accurate test can still be mostly wrong when the base rate is very low.

Question 19

In a SQL query that has both a GROUP BY clause and a window function, in what order do they execute, and what does that imply about which values the window function can see?

Accepted Answer

SQL logical execution order: FROM, WHERE, GROUP BY, HAVING, SELECT (window functions evaluated here, after aggregation), ORDER BY. This means a window function in the SELECT clause sees only the grouped rows, not the original pre-aggregated rows. If you want to apply a window function to unaggregated data, it must be done in a subquery or CTE before the GROUP BY. The candidate should illustrate with an example where a running total over monthly grouped rows behaves differently from one over the raw event rows.

Question 20

Why does the curse of dimensionality make k-nearest neighbors unreliable in high-dimensional spaces, and what does this imply about feature engineering?

Accepted Answer

In high dimensions, all points become approximately equidistant from each other because most volume is near the surface of the hypersphere. This means nearest neighbors are no longer meaningfully close, so k-NN predictions degrade. For feature engineering, this implies that adding irrelevant dimensions hurts distance-based models significantly, so dimensionality reduction (PCA, feature selection) is critical before applying k-NN or density-based clustering. It also explains why deep learning models that learn low-dimensional representations outperform naive methods on raw high-dimensional inputs.

Question 21

How would you use expected value to decide whether to ship a feature that showed a 2% lift in revenue but a 1% lift in support ticket volume? Walk through the calculation.

Accepted Answer

Expected value requires quantifying the dollar value of a 2% revenue lift and the cost of a 1% increase in support tickets. The candidate should walk through a concrete example: if annual revenue is 10M, a 2% lift is 200k. If each support ticket costs 15 dollars and volume is 100k tickets per year, a 1% increase is 1,000 tickets or 15k dollars. Net expected value is 200k minus 15k = 185k, ship it. The candidate should also discuss uncertainty around both estimates and the importance of using confidence intervals rather than point estimates.

Question 22

Explain what the attention mechanism in a transformer does in plain terms, without using the words 'query', 'key', and 'value' as jargon. Why is this better than the fixed-length context vector in RNNs?

Accepted Answer

Attention lets each output token look back at all input tokens and decide which ones are most relevant to it, producing a weighted combination of input representations. This is better than RNNs which compress all input information into a single fixed-size vector and then try to decode from it, which loses information for long sequences. The candidate should explain that attention is O(n^2) in sequence length, which is why it became computationally feasible only as hardware scaled, and why sparse attention mechanisms exist.

Question 23

Explain how MapReduce works conceptually, then write the map and reduce functions to count word frequencies across a large corpus of documents that does not fit in memory on a single machine.

Accepted Answer

A strong answer explains that the map phase produces (key, value) pairs from input chunks in parallel across many workers, and the reduce phase aggregates values for each unique key after a shuffle and sort step. For word count, map emits (word, 1) per word and reduce sums the values for each word. The candidate should mention that the shuffle is the bottleneck, explain how combiners reduce network traffic, and note that MapReduce is largely superseded inside Google by FlumeJava (the inspiration for Apache Beam and Cloud Dataflow), but the concept remains foundational and still appears in interview discussions.

Question 24

Snowflake uses columnar storage optimized for analytics. Explain exactly why columnar storage is faster for aggregate queries but worse for point lookups, how data is compressed in columnar format, and when you would still reach for a traditional row store.

Accepted Answer

Columnar storage keeps all values of a single column contiguous on disk, so a SUM(revenue) query reads only the revenue column, skipping all other columns. This dramatically reduces I/O for wide tables. Compression is much more effective per column because values in a column share the same type and often have low cardinality (run-length encoding, dictionary encoding, delta encoding all apply well). For a point lookup by primary key (fetch one row), you must read from each column file to reconstruct the row, making it slower than a row store where one disk seek gets the full row. Row stores excel at high-concurrency OLTP with frequent single-row reads and writes.

Questions

Applying the CLT to justify using a t-test on non-normal dataTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

Intuition behind why regularization prevents overfittingTechnical fundamentalseasyVery common

As asked

Sample answer outline

Expect these follow-ups

Distinguishing confounding from mediationTechnical fundamentalsmediumOccasional

As asked

Sample answer outline

Expect these follow-ups

When to choose Random Forest over gradient boostingTechnical fundamentalsmediumVery common

As asked

Sample answer outline

Expect these follow-ups

Why ReLU became the default activation in deep networksTechnical fundamentalsmediumOccasional

As asked

Sample answer outline

Expect these follow-ups

How Bayesian updating works in plain termsTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

How SQL window functions execute relative to GROUP BYTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

The curse of dimensionality and its effect on distance-based modelsTechnical fundamentalsmediumOccasional

As asked

Sample answer outline

Expect these follow-ups

Using expected value for product experiment decisionsTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

Intuition behind the attention mechanism in transformersTechnical fundamentalshardOccasional

As asked

Sample answer outline

Expect these follow-ups

Explain MapReduce and implement word countTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

Explain OLAP columnar storage vs OLTP row storage tradeoffsTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

Applying the CLT to justify using a t-test on non-normal dataTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

Intuition behind why regularization prevents overfittingTechnical fundamentalseasyVery common

As asked

Sample answer outline

Expect these follow-ups

Distinguishing confounding from mediationTechnical fundamentalsmediumOccasional

As asked

Sample answer outline

Expect these follow-ups

When to choose Random Forest over gradient boostingTechnical fundamentalsmediumVery common

As asked

Sample answer outline

Expect these follow-ups

Why ReLU became the default activation in deep networksTechnical fundamentalsmediumOccasional

As asked

Sample answer outline

Expect these follow-ups

How Bayesian updating works in plain termsTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

How SQL window functions execute relative to GROUP BYTechnical fundamentalsmediumCommon

As asked

Sample answer outline

Expect these follow-ups

The curse of dimensionality and its effect on distance-based modelsTechnical fundamentalsmediumOccasional

As asked

Sample answer outline