Prepare for and practice interview questions from Dow Jones across topics like Analytics, Statistics, Machine Learning and more.

Dow Jones Interview Questions

Dow Jones Interview Guides

Data Structures & Algorithms

Given two sorted lists, write a function to merge them into one sorted list.

Merge Sorted Lists

Find the missing integer from a array of consequtive integers

Find the Missing Number

Get Top N Frequent Words

String Shift

Given a list of integers, and an integer N, write a function to find all combinations that sum to the value N.

Sum to N

Write a query to select the top 3 departments with at least ten employees and rank them according to the percentage of their employees making over 100K in salary.

Employee Salaries

Select the 2nd highest salary in the engineering department

2nd Highest Salary

Write a SQL query to create a histogram of the number of comments per user in the month of January 2020.

Comments Histogram

Write a query that returns all neighborhoods that have 0 users.

Empty Neighborhoods

Closest SAT Scores

Machine Learning

Encoding Categorical Features

Coefficients of Logistic Regression

Bank Fraud Model

Lasso vs Ridge

Rebalancing outcome probabilities for a classifier on imbalanced data.

Rebalance Probabilities

Probability

500 Cards

Given three uniform(0,4) random variables, what is the probability that the median of them is greater than 3?

Median Probability

Flipping 576 Times

Profit-Maximizing Dice Game

HHT or HTT

Statistics

Write a function to return the cumulative percentage of students that received scores within certain buckets.

Bucket Test Scores

Estimating D

P-value to a Layman

What are the assumptions of linear regression?

Assumptions of Linear Regression

What is the difference between the Z and t tests?

Z and t-Tests

We’re given two tables, a <code>users</code> table with demographic information and the neighborhood they live in and a <code>neighborhoods</code> table.

Write a query that returns all neighborhoods that have 0 users. 

Example:

Input:

<code>users</code> table

<table>
<thead>
<tr>
<th>Columns</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>name</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>neighborhood_id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>
</tbody>
</table>
<code>neighborhoods</code> table

<table>
<thead>
<tr>
<th>Columns</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>name</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>city_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>Columns</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>name</code></td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>

Suppose your company needs to cut costs due to an economic downturn. During a coffee break, you hear a rumor that a lot of money goes to employees who don’t do their work, and you decide to find out if the rumor is true.

Given two tables: <code>employees</code> and <code>projects</code>, find the sum of the salaries of all the employees who didn’t complete any of their projects.

Clarification:

We will consider a project unfinished if it has no end date (its <code>End_dt</code> is <code>NULL</code>).

Given this, we’ll say an employee didn’t finish any of their projects when:

<ul>
<li>They were assigned at least one project.</li>
<li>None of their projects have an <code>End_dt</code>.</li>
</ul>

Example:

Input:

<code>employees</code> table

<table>
<thead>
<tr>
<th>id</th>
<th>salary</th>
</tr>
</thead>

<tbody>
<tr>
<td>INTEGER</td>
<td>FLOAT</td>
</tr>
</tbody>
</table>
<code>projects</code> table

<table>
<thead>
<tr>
<th>employee_id</th>
<th>project_id</th>
<th>Start_dt</th>
<th>End_dt</th>
</tr>
</thead>

<tbody>
<tr>
<td>INTEGER</td>
<td>INTEGER</td>
<td>DATETIME</td>
<td>DATETIME</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>total_slack_salary</th>
</tr>
</thead>

<tbody>
<tr>
<td>INTEGER</td>
</tr>
</tbody>
</table>

Find the total salary of slacking employees.

Slacking Employees Salaries

Write a Python function called <code>max_profit</code> that takes a list of integers, where the i-th integer represents the price of a given stock on day i, and returns the maximum profit you can achieve by buying and selling the stock.

You may complete, at most, two complete buy/sell transactions to maximize profits on a stock.

Input:

<pre tabindex="0" class="chroma"><code>prices = [3, 3, 5, 0, 0, 3, 1, 4]
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def max_profit(prices) -&gt; 6
</code></pre>

Explanation:

<pre><code>Buy on day 4 (price = 0) and sell on day 6 (price = 3): profit = 3 - 0 = 3.
Then buy again on day 7 (price = 1) and sell on day 8 (price = 4): profit = 4 - 1 = 3.
Total profit: 3 + 3 = 6.
</code></pre>

Maximum Profit

You’re a data scientist at Meta working on a feature that helps users discover trending books and understand borrowing behavior across the platform’s digital library.

Every time a user borrows a book, the system logs a <code>borrowed_at</code> timestamp and a <code>returned_at</code> timestamp. The team wants to measure how long books are typically kept, while also flagging which books have large inventories.

Write a SQL query to compute the following metrics for each book:

<ol>
<li><code>avg_borrow_duration_days</code>: average borrowing duration in days (rounded to 2 decimals)</li>
<li><code>has_over_10_copies</code>: <code>&#39;Yes&#39;</code> if the book has more than 10 copies available, otherwise <code>&#39;No&#39;</code></li>
</ol>

Only include completed borrows where <code>returned_at</code> is not NULL.

Sort results by <code>avg_borrow_duration_days</code> DESC, then by <code>book_id</code> ASC.

Note: Every book that has at least one completed borrow should appear in the output.

<h3>Schema</h3>

Input

<code>books</code> Table

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td>book_id</td>
<td>INTEGER</td>
</tr>

<tr>
<td>title</td>
<td>VARCHAR</td>
</tr>

<tr>
<td>copies_available</td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<code>borrows</code> Table

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td>borrow_id</td>
<td>INTEGER</td>
</tr>

<tr>
<td>user_id</td>
<td>INTEGER</td>
</tr>

<tr>
<td>book_id</td>
<td>INTEGER</td>
</tr>

<tr>
<td>borrowed_at</td>
<td>DATETIME</td>
</tr>

<tr>
<td>returned_at</td>
<td>DATETIME</td>
</tr>
</tbody>
</table>
Output

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td>book_id</td>
<td>INTEGER</td>
</tr>

<tr>
<td>avg_borrow_duration_days</td>
<td>DECIMAL</td>
</tr>

<tr>
<td>has_over_10_copies</td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>

<h3>Example</h3>

Input

<code>books</code> Table

<table>
<thead>
<tr>
<th>book_id</th>
<th>title</th>
<th>copies_available</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>Dune</td>
<td>12</td>
</tr>

<tr>
<td>2</td>
<td>Sapiens</td>
<td>8</td>
</tr>

<tr>
<td>3</td>
<td>Clean Code</td>
<td>15</td>
</tr>
</tbody>
</table>
<code>borrows</code> Table

<table>
<thead>
<tr>
<th>borrow_id</th>
<th>user_id</th>
<th>book_id</th>
<th>borrowed_at</th>
<th>returned_at</th>
</tr>
</thead>

<tbody>
<tr>
<td>101</td>
<td>10</td>
<td>1</td>
<td>2026-01-01 09:00:00</td>
<td>2026-01-06 09:00:00</td>
</tr>

<tr>
<td>102</td>
<td>11</td>
<td>1</td>
<td>2026-01-10 10:00:00</td>
<td>2026-01-13 22:00:00</td>
</tr>

<tr>
<td>103</td>
<td>12</td>
<td>2</td>
<td>2026-01-02 12:00:00</td>
<td>2026-01-04 12:00:00</td>
</tr>

<tr>
<td>104</td>
<td>13</td>
<td>2</td>
<td>2026-01-05 09:00:00</td>
<td>NULL</td>
</tr>

<tr>
<td>105</td>
<td>14</td>
<td>3</td>
<td>2026-01-03 08:00:00</td>
<td>2026-01-10 08:00:00</td>
</tr>
</tbody>
</table>
Output

<table>
<thead>
<tr>
<th>book_id</th>
<th>avg_borrow_duration_days</th>
<th>has_over_10_copies</th>
</tr>
</thead>

<tbody>
<tr>
<td>3</td>
<td>7.00</td>
<td>Yes</td>
</tr>

<tr>
<td>1</td>
<td>4.25</td>
<td>Yes</td>
</tr>

<tr>
<td>2</td>
<td>2.00</td>
<td>No</td>
</tr>
</tbody>
</table>
Explanation

For book 1, there are two completed borrows. The durations are 5.00 days (Jan 1 → Jan 6) and 3.50 days (Jan 10 10:00 → Jan 13 22:00). The average borrowing duration is (5.00 + 3.50) / 2 = 4.25 days. Since <code>copies_available = 12</code>, the book is flagged as <code>&#39;Yes&#39;</code>.

For book 2, only one borrow is completed (the other has <code>returned_at = NULL</code> and is excluded). The completed duration is 2.00 days. Since <code>copies_available = 8</code>, it is flagged as <code>&#39;No&#39;</code>.

For book 3, there is one completed borrow lasting 7.00 days, and <code>copies_available = 15</code>, so it is flagged as <code>&#39;Yes&#39;</code>.

Write a SQL query to compute the average borrowing duration and identify books with over 10 copies.

Digital Library Borrowing Metrics

You have a table containing information about the time that employees have worked on a project.

Write a query to find how many unique calendar days each employee worked. Order your query by the <code>employee_id</code>.

Example:

An employee worked on two projects, and recorded their time like this:

<table>
<thead>
<tr>
<th>employee_id</th>
<th>project_id</th>
<th>start_date</th>
<th>end_date</th>
</tr>
</thead>

<tbody>
<tr>
<td>1	</td>
<td>1</td>
<td>2024-01-01</td>
<td>2024-01-06</td>
</tr>

<tr>
<td>1	</td>
<td>2</td>
<td>2024-01-03</td>
<td>2024-01-07</td>
</tr>
</tbody>
</table>
The number of unique calendar days is therefore 7.

<h3>Schema</h3>

Input:

<code>projects</code> table

<table>
<thead>
<tr>
<th>Columns	</th>
<th>	Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>employee_id</code>	</td>
<td>	INTEGER	</td>
</tr>

<tr>
<td><code>project_id</code>	</td>
<td>	INTEGER	</td>
</tr>

<tr>
<td><code>start_date</code>	</td>
<td>	DATE	</td>
</tr>

<tr>
<td><code>end_date</code>	</td>
<td>	DATE	</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>Columns	</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>employee_id</code>	</td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>unique_days_worked</code>	</td>
<td>INTEGER</td>
</tr>
</tbody>
</table>

<h3>Example</h3>

Input:

<code>projects</code> table

<table>
<thead>
<tr>
<th>employee_id</th>
<th>project_id</th>
<th>start_date</th>
<th>end_date</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>1</td>
<td>2024-01-01</td>
<td>2024-01-15</td>
</tr>

<tr>
<td>1</td>
<td>1</td>
<td>2024-01-21</td>
<td>2024-01-31</td>
</tr>

<tr>
<td>1</td>
<td>2</td>
<td>2024-01-25</td>
<td>2024-02-28</td>
</tr>

<tr>
<td>2</td>
<td>3</td>
<td>2024-01-01</td>
<td>2024-01-31</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>employee_id</th>
<th>unique_days_worked</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>54</td>
</tr>

<tr>
<td>2</td>
<td>31</td>
</tr>
</tbody>
</table>

Unique Work Days

Given a sorted list of positive integers with some entries being <code>None</code>, write a function to return a new list where all <code>None</code> values are replaced with the most recent non-<code>None</code> value in the list.

Note: If the first entry in the list is <code>None</code>, assume the previous entry was <code>0</code>.

Example:

Input:

<pre tabindex="0" class="chroma"><code>input_list = [1,2,None,None,4,5,None]
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def fill_none(input_list) -&gt; [1,2,2,2,4,5,5] 
</code></pre>

Write a function to return a new list where all empty values are replaced with the most recent non-empty value in the list.

Fill None Values

You’re given a string that may contain the characters <code>{</code>, <code>}</code>, <code>[</code>, <code>]</code>, <code>(</code>, and <code>)</code>.

Task: Verify that the string is balanced. A balanced string is one where every opening character, <code>{</code>, <code>[</code>, or <code>(</code>, has a corresponding closing character, <code>}</code>, <code>]</code>, or <code>)</code>.

Write a function called <code>is_balanced(string: str) -&gt; bool</code> which verifies the balance of a string.

Example:

<pre tabindex="0" class="chroma"><code>is_balanced(&#39;(())[]{}&#39;) -&gt; True
</code></pre>

<pre tabindex="0" class="chroma"><code>is_balanced(&#39;{([(){}])()}&#39;) -&gt; True
</code></pre>

<pre tabindex="0" class="chroma"><code>is_balanced(&#39;{}[]())&#39;) -&gt; False
</code></pre>

<hr/>

Write a function that tests whether a string of brackets is balanced.

The Brackets Problem

Given a string, find the first non-repeating character in it and return its index. If it doesn’t exist, return -1.

Example questions:

Consider a string where all characters are lowercase alphabets.

Example 1:

Input:

<pre tabindex="0" class="chroma"><code>s = &#39;helloworld&#39;
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>first_uniq_char(s) -&gt; 0
</code></pre>

In the string ‘helloworld’, the first non-repeating character is ‘h’, which is at index 0.

Example 2:

Input:

<pre tabindex="0" class="chroma"><code>s = &#39;concussion&#39;
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>first_uniq_char(s) -&gt; 4
</code></pre>

In the string ‘concussion’, the first non-repeating character is ‘u’, which is at index 4.

This problem involves finding the first non-repeating character in a given string. The solution involves iterating over the string and keeping track of the frequency of each character. The first character that has a frequency of 1 is the first non-repeating character.

Find the First Non-Repeating Character in a String

Let’s say that you’re the PM on Google Maps.

<ol>
<li>How would you improve Google Maps?</li>

<li>What metrics would you check to see if your feature improvements are successful?</li>
</ol>

Google Maps Improvement

Product Sense & Metrics

You are given two rectangles <code>a</code> and <code>b</code> each defined by four ordered pairs denoting their corners on the <code>x</code>, <code>y</code> plane. Write a function <code>rectangle_overlap</code> to determine whether or not they overlap. Return <code>True</code> if so, and <code>False</code> otherwise.

Note: If the two rectangles border one another, or share a corner like two diagonally adjacent positions on a chessboard, they are said to overlap.

Note: The lists of ordered pairs are in no particular order. The first entry in list <code>a</code> could be the top left corner while the first in list <code>b</code> is the bottom right.

Example:

Input:

<pre tabindex="0" class="chroma"><code>
a = [(-3,5), (-3,2),(0,5),(0,2)]
b = [(-1,4), (3,4), (3,1), (-1,1)]
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def rectangle_overlap(a, b) -&gt; True
</code></pre>

As point <code>(0,2)</code> is fully contained in rectangle <code>b</code> and point <code>(-1,4)</code> is fully contained in a rectangle <code>a</code>.

Write a function to determine whether or not two rectangles overlap.

Rectangle Overlap

Let’s say you’re given all the different marketing channels along with their respective marketing costs at a company called Mode, that sells B2B analytics dashboards.

What metrics would you use to determine the value of each marketing channel?

When determining the value of each marketing channel for the company Mode, which metric is considered the key metric?

When determining the value of each marketing channel for the company Mode, which metric is considered the key metric?

Marketing Channel Metrics

What metrics would you use to determine the value of each marketing channel?

Analytics

You’re a data scientist at YouTube. Leadership wants to revisit how the platform measures success, not for a specific new feature, but for the product as a whole given its current state.

How would you define and measure success for YouTube? Identify your primary and secondary metrics, explain your rationale for each, and discuss how you would choose the right thresholds for any engagement-based metrics you propose.

The role of A/B testing in measuring the success rate of an analytics experiment

Success Measurement

Let’s say at Netflix we offer a subscription where customers can enroll for a 30-day free trial. After 30 days, customers will be automatically charged based on the package selected.

Let’s say we want to measure the success of acquiring new users through the free trial.

How can we measure acquisition success and what metrics can we use to measure the success of the free trial?

Let’s say we want to measure the effectiveness of Netflix’s 30-day free trial plan.

What is NOT a metric we would use?

Let's say we want to measure the effectiveness of Netflix's 30-day free trial plan.

What is NOT a metric we would use?

Netflix Retention

How do we measure the success of acquiring new users through a free trial

You are given a binary tree of unique positive numbers. Each node in the binary tree is represented as a dictionary with the following keys:

<ul>
<li><code>data</code>: integer value stored in the node</li>
<li><code>left</code>: reference to the left child (or <code>None</code>)</li>
<li><code>right</code>: reference to the right child (or <code>None</code>)</li>
</ul>

<pre tabindex="0" class="chroma"><code>node = {
 &#34;data&#34;: 6,
 &#34;left&#34;: {
 &#34;data&#34;: 3,
 &#34;left&#34;: {...},
 &#34;right&#34;: {...}
 },
 &#34;right&#34;: {
 &#34;data&#34;: 9,
 &#34;left&#34;: {...},
 &#34;right&#34;: {...}
 }
}
</code></pre>

Given two node values as input (<code>value1</code> and <code>value2</code>), write a function to return the value of the nearest common ancestor (lowest node in the tree that has both nodes as descendants).

Note: If one of the nodes doesn’t exist in the tree, return <code>-1</code>.

Example:

Input:

<pre tabindex="0" class="chroma"><code># Diagram of the binary tree
&#39;&#39;&#39;
 6 
 / \ 
 3 9 
 / \
 2 11
 / \
 5 8
&#39;&#39;&#39;
value1 = 8
value2 = 2
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>common_ancestor(root,value1,value2) -&gt; 3
</code></pre>

Explanation:

<ul>
<li>Ancestors of <code>8</code>: <code>11 → 3 → 6</code></li>
<li>Ancestors of <code>2</code>: <code>3 → 6</code></li>
<li>Common ancestors: <code>3</code> and <code>6</code></li>
<li>The nearest common ancestor is <code>3</code>.</li>
</ul>

Write a function to return the value of the nearest node that is a parent to both nodes.

Nearest Common Ancestor

Given a table of <code>subscriptions</code>, write a query to get the retention rate of each monthly cohort for each <code>plan_id</code> for the three months after sign-up.

Order your output by <code>start_month</code>, <code>plan_id</code>, then <code>num_month</code>.

If an <code>end_date</code> is in the same month as <code>start_date</code> we say the subscription was not retained in the first month.

If the <code>end_date</code> occurs in the month after the month of <code>start_date</code>, the subscription was not retained in the second month. And so on for the third.

The <code>end_date</code> field is <code>NULL</code> if the user has not canceled.

<h3>Schema:</h3>

Input:

<code>subscriptions</code> table

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>user_id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>start_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>end_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>plan_id</code></td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>start_month</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>num_month</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>plan_id</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>retained</code></td>
<td>FLOAT</td>
</tr>
</tbody>
</table>

<h3>Example</h3>

Input:

<code>subscriptions</code> table

<table>
<thead>
<tr>
<th>user_id</th>
<th>start_date</th>
<th>end_date</th>
<th>plan_id</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>2020-01-10 00:00:00</td>
<td>2020-03-31 23:59:59</td>
<td>plan1</td>
</tr>

<tr>
<td>2</td>
<td>2020-01-10 00:00:00</td>
<td>2020-02-01 23:59:59</td>
<td>plan1</td>
</tr>

<tr>
<td>3</td>
<td>2020-01-10 00:00:00</td>
<td>2020-07-31 23:59:59</td>
<td>plan1</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>start_month</th>
<th>num_month</th>
<th>plan_id</th>
<th>retained</th>
</tr>
</thead>

<tbody>
<tr>
<td>2020-01-01</td>
<td>1</td>
<td>plan1</td>
<td>1.0</td>
</tr>

<tr>
<td>2020-01-01</td>
<td>2</td>
<td>plan1</td>
<td>0.67</td>
</tr>

<tr>
<td>2020-01-01</td>
<td>3</td>
<td>plan1</td>
<td>0.33</td>
</tr>
</tbody>
</table>
Explanation:

<ul>
<li>Cohort: All subscriptions starting in January 2020.</li>
<li>Month 1 retention: All 3 users were active in the first month → 3⁄3 = 1.0</li>
<li>Month 2 retention: Only users 1 and 3 were active in Feb → 2⁄3 ≈ 0.67</li>
<li>Month 3 retention: Only user 3 was active in Mar → 1⁄3 ≈ 0.33</li>
</ul>

Subscription Retention

What are the assumptions of linear regression?

Which assumption of the residuals of the standard linear regression model can not be overcome by increasing the sample size?

Regression assumptions

Given a <code>ListNode</code> representing the head of a linked list, write a function <code>is_cyclic</code> which returns <code>True</code> if a linked list is cyclic and <code>False</code> if it is not. A linked list is cyclic when there is no tail to the linked list, and the supposed tail is attached to another node inside the list itself, creating a cycle.

A <code>ListNode</code> is defined as:

<pre tabindex="0" class="chroma"><code>class ListNode:
	value: ListNode = None
 next: ListNode = None
</code></pre>

Example:

Input:

<img src="/s3/questions/1a65651d-a483-4481-b073-2738d65ab9d7.png" alt="image"/>

Output:

<pre tabindex="0" class="chroma"><code>def is_cyclic(head) -&gt; True
</code></pre>

Cyclic Detection

Write a function <code>minimum_change</code> to find the minimum number of coins that make up the given amount of change <code>cents</code>. Assume we only have coins of value 1, 5, 10, and 25 cents.

Example:

Input:

<pre tabindex="0" class="chroma"><code>cents = 73
</code></pre>

Output:

<pre tabindex="0" class="chroma"><code>def minimum_change(cents) -&gt; 7
#(25 + 25 + 10 + 10 + 1 + 1 + 1)
</code></pre>

Minimum Change

Let’s say you’re given a dataframe of standardized test scores from high schoolers from grades 9 to 12 called df_grades.

Given the dataset, write code function in Pandas called bucket_test_scores to return the cumulative percentage of students that received scores within the buckets of &lt;50, &lt;75, &lt;90, &lt;100.

Example:

Input:

<pre tabindex="0" class="chroma"><code>print(df_grades)
</code></pre>

<table>
<thead>
<tr>
<th>user_id</th>
<th>grade</th>
<th>test score</th>
</tr>
</thead>

<tbody>
<tr>
<td>1</td>
<td>10</td>
<td>85</td>
</tr>

<tr>
<td>2</td>
<td>10</td>
<td>60</td>
</tr>

<tr>
<td>3</td>
<td>11</td>
<td>90</td>
</tr>

<tr>
<td>4</td>
<td>10</td>
<td>30</td>
</tr>

<tr>
<td>5</td>
<td>11</td>
<td>99</td>
</tr>
</tbody>
</table>
Output:

<pre tabindex="0" class="chroma"><code>def bucket_test_scores(df_grades) -&gt;
</code></pre>

<table>
<thead>
<tr>
<th>grade</th>
<th>test score</th>
<th>percentage</th>
</tr>
</thead>

<tbody>
<tr>
<td>10</td>
<td>&lt;50</td>
<td>33%</td>
</tr>

<tr>
<td>10</td>
<td>&lt;75</td>
<td>66%</td>
</tr>

<tr>
<td>10</td>
<td>&lt;90</td>
<td>100%</td>
</tr>

<tr>
<td>10</td>
<td>&lt;100</td>
<td>100%</td>
</tr>

<tr>
<td>11</td>
<td>&lt;50</td>
<td>0%</td>
</tr>

<tr>
<td>11</td>
<td>&lt;75</td>
<td>0%</td>
</tr>

<tr>
<td>11</td>
<td>&lt;90</td>
<td>50%</td>
</tr>

<tr>
<td>11</td>
<td>&lt;100</td>
<td>100%</td>
</tr>
</tbody>
</table>

Priority queues are an abstract data structure responsible for allowing enqueuing items with an attached priority. While typically implemented with a heap, implement a priority queue using a linked list.

The Priority Queue implementation should support the following operations:

<ul>
<li><code>insert</code>(element, priority): This operation should be able to insert an element into the Priority Queue, along with its corresponding priority.</li>
<li><code>delete()</code>: This operation should remove and return the element with the highest priority. If multiple elements share the same highest priority, the element first enqueued should be returned. In the case that the queue is empty, return None.</li>
<li><code>peek()</code>: This operation should return the element with the highest priority without removing it from the Priority Queue. Again, in the case of equal highest priorities, the element first enqueued should be returned. In the case that the queue is empty, return None.</li>
</ul>

Note: Smaller priority values imply that they have higher priority.

<h3>Input</h3>

<pre tabindex="0" class="chroma"><code>queue.insert(2, 1)
queue.insert(3, 3)
</code></pre>

<h3>Output:</h3>

<pre tabindex="0" class="chroma"><code>[2,3]
</code></pre>

Implementing a priority queue used linked lists.

Priority Queue Using Linked List

An e-commerce team wants to test whether changing the checkout button color increases conversions. You have an experiment group and a control group, and you need to decide whether the change is worth launching.

How would you set up the A/B test end to end: define the hypothesis, choose the right metrics and guardrails, determine sample size inputs, avoid bias in assignment, decide how long to run the test, and analyze the results before making a launch decision?

An A/B test is being conducted to determine which version of a payment processing page leads to higher conversion rates. You’re responsible for analyzing the results. How would you set up and analyze this A/B test? Additionally, how would you use bootstrap sampling to calculate the confidence intervals for the test results, ensuring your conclusions are statistically valid?

A/B Testing a Checkout Button Change

A/B Testing

You are building a text generation system using a large language model.

During experimentation, you observe that greedy decoding produces repetitive and generic outputs, while unconstrained random sampling often leads to incoherent or low-quality text.

a) Explain how top-k sampling works and how it limits the token selection process.

b) Explain how nucleus (top-p) sampling works and how it differs from top-k sampling.

c) Compare the trade-offs between top-k and top-p sampling in terms of output diversity, stability, and control.

d) Describe a real-world scenario where you would prefer top-k sampling and one where you would prefer nucleus (top-p) sampling.

This question evaluates a candidate’s understanding of probabilistic decoding strategies used in large language models, specifically top-k and nucleus (top-p) sampling. It focuses on conceptual differences, trade-offs, and real-world use cases rather than implementation details.


Top-K vs Nucleus (Top-P) Sampling in LMs

AI & Agentic Systems

Assume you have a logistic model that is heavily weighted on one variable and that one variable has sample data like 50.00, 100.00, 40.00, etc….

Next let’s assume that there was a data quality issue with that variable and an unknown number of values removed the decimal point. For example 100.00 turned into 10000.

Would the model still be valid? Why or why not? How would you fix the model?

In predicting Airbnb booking prices with a dataset containing thousands of diverse features, which regression model could potentially perform better?

In predicting Airbnb booking prices with a dataset containing thousands of diverse features, which regression model could potentially perform better?

Variable Error

Write a SQL query to select the 2nd highest salary in the engineering department.

Note: If more than one person shares the highest salary, the query should select the next highest salary.

Example:

Input:

<code>employees</code> table

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>first_name</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>last_name</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>salary</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>department_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<code>departments</code> table

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>name</code></td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>
Output:

<table>
<thead>
<tr>
<th>Column</th>
<th>Type</th>
</tr>
</thead>

<tbody>
<tr>
<td><code>salary</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>

What’s the difference between Lasso and Ridge Regression?

Under what circumstances is it recommended to use Lasso regression over Ridge regression?

Under what circumstances is it recommended to use Lasso regression over Ridge regression?

Let’s say that you are leading the development of an internal chatbot for operations support at Thomson Reuters, an internal tooling in their news division for referencing articles.

How would you design a system given a choice between fine-tuning the LLM or using Retrieval-Augmented Generation (RAG)?

Fine-Tuning VS RAG

You work as a software engineer at Morningstar. You are tasked with architecting a web-based portfolio management platform where authenticated users can view their investment portfolios, including stocks, mutual funds, and other assets. The system must display real-time asset prices, send notifications for significant price drops or new recommendations from external providers, and handle large volumes of user traffic with scalable caching.

How would you architect this system?

Design a secure, real-time, scalable portfolio management system with notifications and caching.

Portfolio Platform Architecture

Data Pipelines

Let’s say you’re a data engineer at Fidelity Investments, and you’re running a SQL query on a cloud-based data warehouse. All cluster resources and network health metrics look normal, but the query is still taking over 10 minutes to complete.

How would you go about diagnosing and improving the performance of this query?

How would you diagnose and speed up a slow SQL query when system metrics look healthy?

Slow SQL Query

Query Optimization

Let’s say that you’re training a classification model.

How would you combat overfitting when building tree-based models?

Let's say that you're training a classification model.   How would you combat overfitting when building tree-based models?

Overfit Avoidance

You work as a backend engineer at a company deploying a large language model (LLM) as a service for multiple clients. The LLM must handle concurrent user requests while maintaining low latency, high throughput, and robust fault tolerance.

How would you design the serving infrastructure for this LLM to efficiently manage simultaneous requests from multiple users?

Design a system to handle simultaneous requests to a deployed LLM model ensuring scalability, low latency, and reliability.

Calculate Moving Average	SQL	Easy
Predict Customer Churn	Machine Learning	Medium
A/B Test Significance	Statistics	Medium
Optimize Query Performance	SQL	Hard
Feature Importance Analysis	Machine Learning	Medium
Clean Missing Data	Python	Easy
Neural Network Architecture	Deep Learning	Hard
Calculate Cohort Retention	SQL	Medium
Bayesian Probability	Statistics	Easy
Recommend Similar Products	Machine Learning	Hard

Dow Jones Interview Questions

Dow Jones Interview Guides

Dow Jones Interview Questions

Challenge

Dow Jones Salaries by Position

Discussion & Interview Experiences

Discussion & Interview Experiences