Prepare for and practice interview questions from Clarifai.

Clarifai Interview Questions

Clarifai Interview Guides

Data Structures & Algorithms

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

Merge Sorted Lists

Find Bigrams

Write a function to create a single dataframe with complete addresses in the format of street, city, state, zip code.

Complete Addresses

Write a function friendship_timeline to generate an output that lists the pairs of friends with their corresponding timestamps of the friendship beginning and then the timestamp of the friendship ending. 

Friendship Timeline

Write a function n_frequent_words that returns the top N frequent words and their frequencies, and state its run-time

Get Top N Frequent Words

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

Calculate the first touch attribution for each `user_id` that converted.

First Touch Attribution

Write a query to return the two students with the closest test scores and the score difference

Closest SAT Scores

Machine Learning

How would you encode a categorical variable with thousands of distinct values

Encoding Categorical Features

Interpret coefficients of logistic regression for categorical and boolean variables

Coefficients of Logistic Regression

Which model would perform better and why?

Booking Regression

How would you build a job recommendation feed?

Job Recommendation

Write a function calculate_rmse to calculate the root mean squared error from y_pred and y_true

RMS Error

Probability

What is the probability that each subsequent card is larger than the previous drawn card when picking three cards from a shuffled deck of 500 cards

500 Cards

First to Six

What is the probability that it's actually raining in Seattle?

Raining in Seattle

What is the probability that item X would be found on Amazon's website?

Found Item

Compute the expected number of good ads in each scenario and the probability a bad rating came from a lazy rater

Ad Raters

A/B Testing

How would you assess the validity of the result?

Experiment Validity

How do we go about selecting the best 10,000 customers for the pre-launch?

Pre-Launching Shows

How would you determine if the new model predicts delivery times better than the old model

Delivery Estimate Model

Button AB Test

A/B Test Power Size

<ul>
<li>Clarifai is a pioneering full-lifecycle deep learning AI platform specializing in computer vision, natural language processing, and audio recognition. Since its founding in 2013 by Matt Zeiler, Ph.D., the company has led the AI industry, transforming unstructured data into structured insights far quicker and more accurately than humans can manage. With a distributed team across the U.S., Canada, Argentina, India, and Estonia, Clarifai is driven by innovation and excellence.</li>
</ul>
<p>Backed by over $100M in funding from notable investors like Menlo Ventures and Google Ventures, Clarifai continues to shape the future of AI. As we delve into this guide, you'll discover our interview process, common questions, and the opportunity to significantly impact both Clarifai and the broader AI landscape. Don't miss this chance to be part of something extraordinary!</p>

- Clarifai is a pioneering full-lifecycle deep learning AI platform specializing in computer vision, natural language processing, and audio recognition. Since its founding in 2013 by Matt Zeiler, Ph.D., the company has led the AI industry, transforming unstructured data into structured insights far quicker and more accurately than humans can manage. With a distributed team across the U.S., Canada, Argentina, India, and Estonia, Clarifai is driven by innovation and excellence.

Backed by over $100M in funding from notable investors like Menlo Ventures and Google Ventures, Clarifai continues to shape the future of AI. As we delve into this guide, you'll discover our interview process, common questions, and the opportunity to significantly impact both Clarifai and the broader AI landscape. Don't miss this chance to be part of something extraordinary!

Overview

<p>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.</p>

<p>Write a query that returns all neighborhoods that have 0 users. </p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>users</code> table</p>

<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>
<p><code>neighborhoods</code> table</p>

<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>
<p><strong>Output:</strong></p>

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

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


<p>Write a SQL query to select the 2nd highest salary in the engineering department.</p>

<p><strong>Note:</strong> If more than one person shares the highest salary, the query should select the next highest salary.</p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>employees</code> table</p>

<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>
<p><code>departments</code> table</p>

<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>
<p><strong>Output:</strong></p>

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

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


<p>Let’s say that your company is running a standard control and variant AB test on a feature to increase conversion rates on the landing page. The PM checks the results and finds a .04 p-value.</p>

<p>How would you assess the validity of the result?</p>


<p>Given the <code>employees</code> and <code>departments</code> table, write a query to get the top 3 highest employee salaries by department. If the department contains less that 3 employees, the top 2 or the top 1 highest salaries should be listed (assume that each department has at least 1 employee). </p>

<p><em>Note: The output should include the full name of the employee in one column, the department name, and the salary. The output should be sorted by department name in ascending order and salary in descending order.</em> </p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>employees</code> table</p>

<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>
<p><code>departments</code> table</p>

<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>
<p><strong>Output</strong>:</p>

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

<tbody>
<tr>
<td><code>employee_nam</code>e</td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>department_name</code></td>
<td>VARCHAR</td>
</tr>

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


Get the top 3 highest employee salaries by department

Top Three Salaries

<p>Amy and Brad take turns in rolling a fair six-sided die. Whoever rolls a “6” first wins the game. Amy starts by rolling first.</p>

<p>What’s the probability that Amy wins?</p>


Amy and Brad take turns in rolling a fair six-sided die. Whoever rolls a “6” first wins the game. Amy starts by rolling first.

What’s the probability that Amy wins?

<p>Given two sorted lists, write a function to merge them into one sorted list.</p>

<p><em>Bonus: What’s the time complexity?</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<pre tabindex="0" class="chroma"><code><span class="line"><span class="cl"><span class="n">list1</span> <span class="o">=</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">5</span><span class="p">]</span>
</span></span><span class="line"><span class="cl"><span class="n">list2</span> <span class="o">=</span> <span class="p">[</span><span class="mi">2</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">6</span><span class="p">]</span>
</span></span></code></pre>

<p><strong>Output:</strong></p>

<pre tabindex="0" class="chroma"><code><span class="line"><span class="cl"><span class="k">def</span> <span class="nf">merge_list</span><span class="p">(</span><span class="n">list1</span><span class="p">,</span><span class="n">list2</span><span class="p">)</span> <span class="o">-&gt;</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">,</span><span class="mi">6</span><span class="p">]</span>
</span></span></code></pre>


<p>You have two sorted lists, one with m elements and one with n elements. What would be the time complexity of merging the two lists via merge sort?</p>


You have two sorted lists, one with m elements and one with n elements. What would be the time complexity of merging the two lists via merge sort? 

Merging Lists Complexity

<p>We’re given a table of bank transactions with three columns, <code>user_id</code>, a deposit or withdrawal value (determined if the value is positive or negative), and <code>created_at</code> time for each transaction.</p>

<p>Write a query to get the total three-day rolling average for deposits by day.</p>

<p><em>Note: Please use the format <code>&#39;%Y-%m-%d&#39;</code> for the date in the outout</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>bank_transactions</code> table</p>

<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>created_at</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>transaction_value</code></td>
<td>FLOAT</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

<tbody>
<tr>
<td><code>dt</code></td>
<td>VARCHAR</td>
</tr>

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


Write a query to get the total three-day rolling average for deposits by day

Rolling Bank Transactions

<p>Write a query to identify customers who placed more than three transactions each in both 2019 and 2020.</p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>transactions</code> table</p>

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

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>product_id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>quantity</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><code>users</code> table</p>

<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>
<p> 
<strong>Output:</strong></p>

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

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


<p>Which of the following queries will return a list of customers and the total amount they have spent on orders?</p>


Which of the following queries will return a list of customers and the total amount they have spent on orders?

Customer Orders

Write a query to identify customers who placed more than three transactions each in both 2019 and 2020

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

<p><em>Note: Assume bin buckets class intervals of one.</em></p>

<p><em>Note: Comments that were created outside of January 2020 should be counted in a “0” bucket</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>users</code> table</p>

<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>created_at</code></td>
<td>DATETIME</td>
</tr>

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

<tr>
<td><code>mail</code></td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>
<p><code>comments</code> table</p>

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

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

<tr>
<td><code>body</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

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

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


<p>Let’s say we’re comparing two machine learning algorithms. In which case would you use a bagging algorithm versus a boosting algorithm? </p>

<p>Give an example of the tradeoffs between the two.</p>


<p>In machine learning, when would you use a bagging algorithm over a boosting algorithm?</p>


In machine learning, when would you use a bagging algorithm over a boosting algorithm?

Bagging vs. Boosting

In which case would you use a bagging algorithm versus a boosting algorithm

Bagging vs Boosting

<p>Given a table of students and their SAT test scores, write a query to return the two students with the closest test scores with the score difference.</p>

<p>If there are multiple students with the same minimum score difference, select the student name combination that is higher in the alphabet. </p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>scores</code> table</p>

<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>student</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>score</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

<tbody>
<tr>
<td><code>one_student</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>other_student</code></td>
<td>VARCHAR</td>
</tr>

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


<p>Imagine a deck of 500 cards numbered from 1 to 500. If all the cards are shuffled randomly and you are asked to pick three cards, one at a time, what’s the probability of each subsequent card being larger than the previous drawn card?</p>


<p>Imagine a deck of 2000 cards numbered from 1 to 2000. If all the cards are shuffled randomly and you are asked to pick three cards, one at a time, what’s the probability of each subsequent card being larger than the previously drawn card?</p>


Imagine a deck of 2000 cards numbered from 1 to 2000. If all the cards are shuffled randomly and you are asked to pick three cards, one at a time, what's the probability of each subsequent card being larger than the previously drawn card?

2000 cards

<p>The schema below is for a retail online shopping company consisting of two tables, <code>attribution</code> and <code>user_sessions</code>.</p>

<ul>
<li><p>The attribution table logs a session visit for each row.</p></li>

<li><p>If <code>conversion</code> is <code>true</code>, then the user converted to buying on that session.</p></li>

<li><p>The <code>channel</code> column represents which advertising platform the user was attributed to for that specific session.</p></li>

<li><p>Lastly the <code>user_sessions</code> table maps many to one session visits back to one user.</p></li>
</ul>

<p>First touch attribution is defined as the channel with which the converted user was associated when they first discovered the website.</p>

<p>Calculate the first touch attribution for each <code>user_id</code> that converted. </p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>attribution</code> table</p>

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

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

<tr>
<td><code>channel</code></td>
<td>VARCHAR</td>
</tr>

<tr>
<td><code>conversion</code></td>
<td>BOOLEAN</td>
</tr>
</tbody>
</table>
<p><code>user_sessions</code> table</p>

<table>
<thead>
<tr>
<th>column</th>
<th>type</th>
</tr>
</thead>

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

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>user_id</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><strong>Example output:</strong></p>

<table>
<thead>
<tr>
<th><code>user_id</code></th>
<th><code>channel</code></th>
</tr>
</thead>

<tbody>
<tr>
<td>123</td>
<td>facebook</td>
</tr>

<tr>
<td>145</td>
<td>google</td>
</tr>

<tr>
<td>153</td>
<td>facebook</td>
</tr>

<tr>
<td>172</td>
<td>organic</td>
</tr>

<tr>
<td>173</td>
<td>email</td>
</tr>
</tbody>
</table>


<p>Given a table of product subscriptions with a subscription start date and end date for each user, write a query that returns true or false whether or not each user has a subscription date range that overlaps with any other completed subscription.</p>

<p>Completed subscriptions have <code>end_date</code> recorded.</p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>subscriptions</code> table</p>

<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>
</tbody>
</table>

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

<tbody>
<tr>
<td>1</td>
<td>2019-01-01</td>
<td>2019-01-31</td>
</tr>

<tr>
<td>2</td>
<td>2019-01-15</td>
<td>2019-01-17</td>
</tr>

<tr>
<td>3</td>
<td>2019-01-29</td>
<td>2019-02-04</td>
</tr>

<tr>
<td>4</td>
<td>2019-02-05</td>
<td>2019-02-10</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

<table>
<thead>
<tr>
<th><code>user_id</code></th>
<th><code>overlap</code></th>
</tr>
</thead>

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

<tr>
<td>2</td>
<td>1</td>
</tr>

<tr>
<td>3</td>
<td>1</td>
</tr>

<tr>
<td>4</td>
<td>0</td>
</tr>
</tbody>
</table>


Write a query to return whether each user's subscription date range overlaps with any other completed subscription

Subscription Overlap

<p>Netflix has hired people to rate movies.</p>

<p>Out of all of the raters, 80% of the raters carefully rate movies and rate 60% of the movies as good and 40% as bad. The other 20% are lazy raters and rate 100% of the movies as good.</p>

<p>Assuming all raters rate the same amount of movies, what is the probability that a movie is rated good?</p>


<p>There are two categories of movie raters: careful and lazy. Careful raters make up 80% of the total. They rate 60% of the movies as good and 40% as bad. Lazy raters, constitute 20% of the raters and 100% of the movies as good. Given these parameters, calculate the overall probability of a movie being rated good.</p>

There are two categories of movie raters: careful and lazy. Careful raters make up 80% of the total. They rate 60% of the movies as good and 40% as bad. Lazy raters, constitute 20% of the raters and 100% of the movies as good. Given these parameters, calculate the overall probability of a movie being rated good.

Lazy Raters

What is the probability that a movie is rated good

<p>We’re given a table of product purchases. Each row in the table represents an individual user product purchase.</p>

<p>Write a query to get the number of customers that were upsold, or in other words, the number of users who bought additional products after their first purchase.</p>

<p><em>Note: If the customer purchased two things on the same day, that does not count as an upsell, as they were purchased within a similar timeframe.</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>transactions</code> table</p>

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

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>product_id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>quantity</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

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


Write a query to get the number of customers that were upsold

Upsell Transactions

<p>Write a query to show the number of users, number of transactions placed, and total order amount per month in the year 2020. Assume that we are only interested in the monthly reports for a single year (January-December).</p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>transactions</code> table</p>

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

<tr>
<td><code>created_at</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>product_id</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>quantity</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><code>products</code> table</p>

<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>

<tr>
<td><code>price</code></td>
<td>FLOAT</td>
</tr>
</tbody>
</table>
<p><code>users</code> table</p>

<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>

<tr>
<td><code>sex</code></td>
<td>VARCHAR</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

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

<tr>
<td><code>num_customers</code></td>
<td>INTEGER</td>
</tr>

<tr>
<td><code>num_orders</code></td>
<td>INTEGER</td>
</tr>

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


Write a query to show the number of users, transactions, and total order amount per month in 2020

Monthly Customer Report

<p>A team wants to A/B test multiple different changes through a sign-up funnel.</p>

<p>For example, on a page, a button is currently red and at the top of the page. They want to see if changing a button from red to blue and/or from the top of the page to the bottom of the page will increase click-through.</p>

<p>How would you set up this test?</p>


<p>Your team wants to run an AB test on the color (red or blue) and position (top/bottom of the page) of a button that links to a promotion.</p>

<p>How many different variants of the test should you run to see the effect that a red button has on the bottom of the page, if currently the button is blue and on the top of the page.</p>


Your team wants to run an AB test on the color (red or blue) and position (top/bottom of the page) of a button that links to a promotion.

How many different variants of the test should you run to see the effect that a red button has on the bottom of the page, if currently the button is blue and on the top of the page.

A/B Test Variants

<p>You are about to get on a plane to Seattle. You want to know if you should bring an umbrella. You call 3 random friends of yours who live there and ask each independently if it’s raining. Each of your friends has a <sup>2</sup>⁄<sub>3</sub> chance of telling you the truth and a <sup>1</sup>⁄<sub>3</sub> chance of messing with you by lying. All 3 friends tell you that “Yes” it is raining.</p>

<p>What is the probability that it’s actually raining in Seattle?</p>


<p>What is the <strong>main</strong> difference between the Frequentist and Bayesian methods of probabilistic analysis?</p>


What is the **main** difference between the Frequentist and Bayesian methods of probabilistic analysis?

Frequentist Vs Bayesian Analysis

<p>Let’s say you want to test the close friends feature on Instagram Stories.</p>

<p>How would you make a control group and test group to account for network effects?</p>


<p>What could be a potential risk when Facebook segments the metrics of the user interface change by market or demographic groups?</p>

What could be a potential risk when Facebook segments the metrics of the user interface change by market or demographic groups?

Network Experiment Design I

<p>What is the primary metric that Facebook should monitor if it decides to make the user interface of its posting feature more like Instagram's?</p>

What is the primary metric that Facebook should monitor if it decides to make the user interface of its posting feature more like Instagram's?

Network Experiment Design II

<p>If Facebook changes its user interface to mimic Instagram's, which adverse effect should they anticipate and monitor?</p>

If Facebook changes its user interface to mimic Instagram's, which adverse effect should they anticipate and monitor?

Network Experiment Design III

How would you make a control group and test group to account for network effects

Network Experiment Design

<p>Let’s say that you’re working on a job recommendation engine. You have access to all user Linkedin profiles, a list of jobs each user applied to, and answers to questions that the user filled in about their job search.</p>

<p>Using this information, how would you build a job recommendation feed?</p>


<p>What problem can arise in job recommendation systems with regard to initial recommendations?</p>

What problem can arise in job recommendation systems with regard to initial recommendations?

<p>Let’s say we have a very naive advertising platform. There is an audience of size <em>A</em> and a limited amount of impressions <em>B</em>.</p>

<p>Each of the impressions goes to just one user, defined at random. Every user in the audience has the same random chance of receiving each impression. </p>

<ol>
<li><p>Compute the probability that a user has exactly 0 impressions. </p></li>

<li><p>What’s the probability that every person has at least 1 impression?</p></li>
</ol>


Compute the probability a user gets exactly 0 impressions, and the probability everyone gets at least 1 impression

Impression Reach

<p>Let’s say that you work at a bank that wants to build a model to detect fraud on the platform.</p>

<p>The bank wants to implement a text messaging service in addition that will text customers when the model detects a fraudulent transaction in order for the customer to approve or deny the transaction with a text response.</p>

<p>How would we build this model?</p>


<p>Let’s say that you work at a bank that wants to build a model to detect fraud on the platform.</p>

<p>The bank wants to implement a text messaging service that will text customers when the model detects a fraudulent transaction in order for the customer to approve or deny the transaction with a text response.</p>

<p>Which statement is true?</p>


Let's say that you work at a bank that wants to build a model to detect fraud on the platform.

The bank wants to implement a text messaging service that will text customers when the model detects a fraudulent transaction in order for the customer to approve or deny the transaction with a text response.

Which statement is true?

Bank Fraud Model

How would we build a model to detect fraud and text customers to approve or deny fraudulent transactions

<p>Let’s say you work at a file-hosting website. You have information on user’s daily downloads in the <code>download_facts</code> table</p>

<p>Use the window function <code>RANK</code> to display the top three users by downloads each day. Order your data by <code>date</code>, and then by <code>daily_rank</code></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>download_facts</code> table</p>

<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>date</code></td>
<td>DATE</td>
</tr>

<tr>
<td><code>downloads</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

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

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

<tr>
<td><code>date</code></td>
<td>DATE</td>
</tr>

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


Write a query to display the top three users by downloads each day

Top 3 Users

<p>Write a function <code>compute_deviation</code> that takes in a list of dictionaries with a key and list of integers and returns a dictionary with the standard deviation of each list.</p>

<p><em>Note: This should be done without using the NumPy built-in functions.</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<pre tabindex="0" class="chroma"><code><span class="line"><span class="cl"><span class="nb">input</span> <span class="o">=</span> <span class="p">[</span>
</span></span><span class="line"><span class="cl">    <span class="p">{</span>
</span></span><span class="line"><span class="cl">        <span class="s1">&#39;key&#39;</span><span class="p">:</span> <span class="s1">&#39;list1&#39;</span><span class="p">,</span>
</span></span><span class="line"><span class="cl">        <span class="s1">&#39;values&#39;</span><span class="p">:</span> <span class="p">[</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">4</span><span class="p">,</span><span class="mi">5</span><span class="p">,</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">],</span>
</span></span><span class="line"><span class="cl">    <span class="p">},</span>
</span></span><span class="line"><span class="cl">    <span class="p">{</span>
</span></span><span class="line"><span class="cl">        <span class="s1">&#39;key&#39;</span><span class="p">:</span> <span class="s1">&#39;list2&#39;</span><span class="p">,</span>
</span></span><span class="line"><span class="cl">        <span class="s1">&#39;values&#39;</span><span class="p">:</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span><span class="mi">1</span><span class="p">,</span><span class="mi">34</span><span class="p">,</span><span class="mi">12</span><span class="p">,</span><span class="mi">40</span><span class="p">,</span><span class="mi">3</span><span class="p">,</span><span class="mi">9</span><span class="p">,</span><span class="mi">7</span><span class="p">],</span>
</span></span><span class="line"><span class="cl">    <span class="p">}</span>
</span></span><span class="line"><span class="cl"><span class="p">]</span>
</span></span></code></pre>

<p><strong>Output:</strong></p>

<pre tabindex="0" class="chroma"><code><span class="line"><span class="cl"> <span class="n">output</span> <span class="o">=</span> <span class="p">{</span><span class="s1">&#39;list1&#39;</span><span class="p">:</span> <span class="mf">1.12</span><span class="p">,</span> <span class="s1">&#39;list2&#39;</span><span class="p">:</span> <span class="mf">14.19</span><span class="p">}</span>
</span></span></code></pre>


Write a function compute_deviation that returns the standard deviation of each list in a list of dictionaries

Compute Deviation

Statistics

<p>A jar holds 1000 coins. Out of all of the coins, 999 are fair and one is double-sided with two heads. Picking a coin at random, you toss the coin ten times.</p>

<p>Given that you see 10 heads, what is the probability that the coin is double headed and the probability that the next toss of the coin is also a head?</p>


Compute the probability the coin is double headed and the probability the next toss is a head given 10 heads

Jars and Coins

<p>Given two strings <code>A</code> and <code>B</code>, write a function <code>can_shift</code> to return whether or not <code>A</code> can be shifted some number of places to get <code>B</code>.</p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<pre tabindex="0" class="chroma"><code><span class="line"><span class="cl"><span class="n">A</span> <span class="o">=</span> <span class="s1">&#39;abcde&#39;</span>
</span></span><span class="line"><span class="cl"><span class="n">B</span> <span class="o">=</span> <span class="s1">&#39;cdeab&#39;</span>
</span></span><span class="line"><span class="cl"><span class="n">can_shift</span><span class="p">(</span><span class="n">A</span><span class="p">,</span> <span class="n">B</span><span class="p">)</span> <span class="o">==</span> <span class="kc">True</span>
</span></span><span class="line"><span class="cl">
</span></span><span class="line"><span class="cl"><span class="n">A</span> <span class="o">=</span> <span class="s1">&#39;abc&#39;</span>
</span></span><span class="line"><span class="cl"><span class="n">B</span> <span class="o">=</span> <span class="s1">&#39;acb&#39;</span>
</span></span><span class="line"><span class="cl"><span class="n">can_shift</span><span class="p">(</span><span class="n">A</span><span class="p">,</span> <span class="n">B</span><span class="p">)</span> <span class="o">==</span> <span class="kc">False</span>
</span></span></code></pre>


Write a function can_shift to return whether or not A can be shifted some number of places to get B

String Shift

<p>Given two tables:  <code>accounts</code>, and <code>downloads</code>, find the average number of downloads for free vs paying accounts, broken down by day.</p>

<p><em>Note: You only need to consider accounts that have had at least one download before when calculating the average.</em></p>

<p><em>Note: round <code>average_downloads</code> to 2 decimal places.</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>accounts</code> table</p>

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

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

<tr>
<td><code>paying_customer</code></td>
<td>BOOLEAN</td>
</tr>
</tbody>
</table>
<p><code>downloads</code> table</p>

<table>
<thead>
<tr>
<th>column</th>
<th>type</th>
</tr>
</thead>

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

<tr>
<td><code>download_date</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>downloads</code></td>
<td>INTEGER</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

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

<tr>
<td><code>paying_customer</code></td>
<td>BOOLEAN</td>
</tr>

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


Write a query to find the average downloads for free vs paying accounts by day, rounded to 2 decimals

Download Facts

<p>Given a table of bank transactions with columns <code>id</code>, <code>transaction_value</code>, and <code>created_at</code> representing the date and time for each transaction, write a query to get the last transaction for each day.</p>

<p>The output should include the id of the transaction, datetime of the transaction, and the transaction amount. Order the transactions by datetime.</p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<p><code>bank_transactions</code> table</p>

<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>created_at</code></td>
<td>DATETIME</td>
</tr>

<tr>
<td><code>transaction_value</code></td>
<td>FLOAT</td>
</tr>
</tbody>
</table>
<p><strong>Output:</strong></p>

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

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

<tr>
<td><code>transaction_value</code></td>
<td>FLOAT</td>
</tr>

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


Last Transaction

<p>Let’s say you work at Allstate. Allstate is running <code>N</code>  online ads right now. The table <code>ads</code> contains all those ads, ranked by popularity via the <code>id</code> column (e.g., the entry with <code>id = 1</code> is the most popular, etc.).</p>

<p>Create a subquery or common table expression named <code>top_ads</code> containing the top 3 ads (by popularity) and return the number of rows that would result from the following operations</p>

<ol>
<li><code>ads INNER JOIN top_ads</code></li>
<li><code>ads LEFT JOIN top_ads</code></li>
<li><code>ads RIGHT JOIN top_ads</code></li>
<li><code>ads CROSS JOIN top_ads</code></li>
</ol>

<p><em>Note: Please make the <code>join_type</code> column in your output have the values <code>inner_join</code>, <code>left_join</code>, etc. for each of their respective join types</em></p>

<p><em>Note: Please return only one query with each number in a different row</em></p>

<p><strong>Example:</strong></p>

<p><strong>Input:</strong></p>

<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>
<p><strong>Output:</strong></p>

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

<tbody>
<tr>
<td><code>join_type</code></td>
<td>VARCHAR</td>
</tr>

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


Create top_ads with the top 3 ads and return the row counts for inner, left, right, and cross joins with ads

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

Clarifai Interview Questions

Clarifai Interview Guides

Overview

Clarifai Interview Questions

Challenge

Discussion & Interview Experiences

Discussion & Interview Experiences