Machine learning software engineer interviews at Google are really challenging. The questions are difficult, specific to Google, and they cover a wide range of topics.
The good news is that the right preparation can make a big difference and can help you land an ML job at Google. We have put together the ultimate guide below, to help you maximize your chances of success.
Here's an overview of what we'll cover:
- Process and timeline
- Example questions
- Preparation tips
What's the Google machine learning engineer interview process and timeline? It normally follows the steps below and takes about two to three months to complete:
1.1 What interviews to expect
- Resume screen
- Recruiter screen (~30 min)
- Technical phone screen (one or two interviews, 45-60 min each)
- Onsite interviews (four to six rounds, 45-60 min each)
Next, we'll dig into each of these steps in more detail. If you're interviewing with multiple companies, take a look at our guides to the Facebook ML engineer interview and the Amazon ML engineer interview.
1.1.1 Resume screen
First, recruiters will look at your resume and assess if your experience matches the open position. This is the most competitive step in the process, as millions of candidates do not make it past this stage. Check out our free Google resume guide with examples for help on writing yours.
So take extra care to tailor your resume to the specific position you're applying to.
If you’re looking for expert feedback, get input from our team of ex-FAANG recruiters, who will cover what achievements to focus on (or ignore), how to fine tune your bullet points, and more.
1.1.2 Recruiter phone screen
In most cases, you'll start your interview process with Google by talking to an HR recruiter on the phone. They are looking to confirm that you've got a chance of getting the job at all, so be prepared to explain your background and why you’re a good fit at Google. You should expect typical behavioral and resume questions, like "tell me about yourself," "why Google?" or "tell me about your current project."
If you get past this first HR screen, the recruiter will then help schedule your technical phone screen(s). Your recruiter will usually let you know who you are interviewing with and what type of interviews you should expect, and they will share resources to help you prepare for them.
1.1.3 Technical phone screen(s)
Next, you'll have one or two problem solving/coding interviews, during which you'll be asked primarily data structure and algorithm questions. These questions tend to be quite similar to the questions you'd encounter as a Google software engineer.
Your interviewer may start with a few behavioral questions, but most of the time will be spent on coding questions. You'll share a Google Doc with your interviewer, write your solution directly in the document, and won't have access to syntax highlighting or auto-completion like you would in a regular IDE. It's therefore a good idea to practice writing code in Google Docs before your interview.
1.1.4 Onsite interviews
Finally, if you pass your technical screens, you’ll be invited to the onsite interviews. These are the real test. You'll typically spend a full day at a Google office and do four to six interviews in total. Each interview will last about 45 minutes and cover one of the following topics:
- Coding interview, where you'll solve algorithm and data structure questions similar to those you'd encounter in a software engineer interview.
- System design interview (for candidates who are Level 5 and above) where you'll be asked to create a high-level design for a modern technology system like Google Cloud, Gmail, etc.
- Machine learning design interview, where you'll need to suggest an approach for how to solve a problem using a machine learning solution.
- Behavioral interview, where you can expect questions about your background, accomplishments, and your motivation for applying to Google.
While you’re onsite, you’ll typically have one or two coding interviews, one system design interview, one or two machine learning interviews, and a behavioral round. However, the exact breakdown will depend on the exact team and role that you’re applying for. For more information about the Google system design interview, take a look at our complete guide.
[COVID note] It's likely that your onsite interviews will be held virtually instead of in-person, due to COVID-19 protocols. However, your recruiter should be able to provide you with the most up-to-date information on Google's onsite interview procedures. Feel free to ask them for details after you've been officially invited to participate in the onsite interviews.
1.2 What exactly is Google looking for?
At the end of each interview, your interviewer will grade your performance using a standardized feedback form that summarizes the attributes Google looks for in a candidate. That form is constantly evolving, but we have listed the main components we know of at the time of writing this article below.
A) Questions asked
In the first section of the form, the interviewer fills in the questions they asked you. These questions are then shared with your future interviewers so you don't get asked the same questions twice.
B) Attribute scoring
Each interviewer will assess you on the four main attributes Google looks for when hiring:
- General cognitive ability. This is often referred to as "GCA" by Googlers. The company wants to hire smart engineers who can learn and adapt to new situations. Here, your interviewer will try to understand how you solve hard problems and how you learn. For more information, take a look at our guide to the Google GCA interview.
- Role-related knowledge and experience. This is often referred to as "RRK" or "RRKE" internally. The company wants to make sure that you have the right experience, domain expertise, and competencies for the position you're applying for. For more information, take a look at our guide to the Google RRK interview.
- Leadership. Google looks for a particular type of leadership called “emergent leadership.” You'll typically be working in cross-functional teams at Google, and different team members are expected to step up and lead at different times in the lifecycle of a project when their skills are needed.
- Googleyness (i.e. culture fit). The company wants to make sure Google is the right environment for you. Your interviewer will check whether you naturally exhibit the company's values, including being comfortable with ambiguity, having a bias to action, and a collaborative nature.
Depending on the exact job you're applying for, these attributes might be broken down further. For instance, "role-related knowledge and experience" could be broken down into "deep learning" or “natural language processing" for certain roles. But the total number of attributes does not usually exceed six or seven.
In this middle section, Google's interviewers typically repeat the questions they asked you, document your answers in detail, and give you a score for each attribute (e.g. "Poor," "Mixed," "Good," "Excellent").
C) Final recommendation
Finally, interviewers will write a summary of your performance and provide an overall recommendation on whether they think Google should be hiring you or not (e.g. "Strong no hire," "No hire," "Leaning no hire," "Leaning hire," "Hire," "Strong hire").
1.3 What happens behind the scenes
If things go well at your onsite interviews, here is what the final steps of the process look like:
- Interviewers submit feedback
- Hiring committee recommendation
- Team matching
- Senior leader and compensation committee review
- Final executive review (only senior roles)
- You get an offer
After your onsite, your interviewers will all submit their feedback usually within two to three days. This feedback will then be reviewed by a hiring committee, along with your resume, internal referrals, and any past work you have submitted. At this stage, the hiring committee will make a recommendation on whether Google should hire you or not.
If the committee recommends that you get hired, you'll usually start your team matching process. In other words, you'll talk to hiring managers and one or several of them will need to be willing to take you in their team in order for you to get an offer from the company.
In parallel, the hiring committee recommendation will be reviewed and validated by a senior manager and a compensation committee who will decide how much money you are offered. Finally, if you are interviewing for a senior role, a senior Google executive will review a summary of your candidacy and compensation before the offer is sent to you.
As you've probably gathered by now, Google goes to great lengths to avoid hiring the wrong candidates. This hiring process with multiple levels of validations helps them scale their teams while maintaining a high caliber of employee. But it also means that the typical process can spread over multiple months.
Okay, now that we've covered the interview process, let's dig into the four types of interviews that you'll encounter:
- System design
- Machine learning design
Below, we've put together a summarized list of example questions for each of these interview types. We've also provided a few notes about them, which should help you get a better idea of what to expect.
Google’s machine learning software engineers solve some of the most difficult problems the company faces with code. It's therefore essential that they have strong problem solving skills. This is the part of the interview where you want to show that you think in a structured way and write code that's accurate, bug-free, and fast.
Here are the most common question types asked in Google coding interviews and their frequency. The percentages and examples come from Glassdoor data on Google software engineers but apply to machine learning candidates as well.
- Graphs / Trees (39% of questions, most frequent)
- Arrays / Strings (26%)
- Dynamic programming (12%)
- Recursion (12%)
- Geometry / Maths (11% of questions, least frequent)
Below, we've listed common examples used during Google software engineer interviews for each of these different question types. To make these questions easier to study, we've modified the phrasing to match the closest problem on Leetcode or another resource, and we've linked to a free solution.
Example Google machine learning engineer interview questions: Coding
- "Given a binary tree, find the maximum path sum. The path may start and end at any node in the tree." (Solution)
- "Given an encoded string, return its decoded string." (Solution)
- "We can rotate digits by 180 degrees to form new digits. When 0, 1, 6, 8, 9 are rotated 180 degrees, they become 0, 1, 9, 8, 6 respectively. When 2, 3, 4, 5, and 7 are rotated 180 degrees, they become invalid. A confusing number is a number that when rotated 180 degrees becomes a different number with each digit valid.(Note that the rotated number can be greater than the original number.) Given a positive integer N, return the number of confusing numbers between 1 and N inclusive." (Solution)
- "Given two words (beginWord and endWord), and a dictionary's word list, find the length of the shortest transformation sequence from beginWord to endWord, such that: 1) Only one letter can be changed at a time, and 2) Each transformed word must exist in the word list." (Solution)
- "Given a matrix of N rows and M columns. From m[i][j], we can move to m[i+1][j], if m[i+1][j] > m[i][j], or can move to m[i][j+1] if m[i][j+1] > m[i][j]. The task is print longest path length if we start from (0, 0)." (Solution)
- Implement a SnapshotArray that supports pre-defined interfaces (note: see link for more details). (Solution)
- "In a row of dominoes, A[i] and B[i] represent the top and bottom halves of the i-th domino. (A domino is a tile with two numbers from 1 to 6 - one on each half of the tile.) We may rotate the i-th domino, so that A[i] and B[i] swap values. Return the minimum number of rotations so that all the values in A are the same, or all the values in B are the same. If it cannot be done, return -1." (Solution)
- "Your friend is typing his name into a keyboard. Sometimes, when typing a character c, the key might get long pressed, and the character will be typed 1 or more times. You examine the typed characters of the keyboard. Return True if it is possible that it was your friends name, with some characters (possibly none) being long pressed." (Solution)
- "Given a string S and a string T, find the minimum window in S which will contain all the characters in T in complexity O(n)." (Solution)
- "Given a list of query words, return the number of words that are stretchy." Note: see link for more details. (Solution)
- Dynamic Programming (12%)
- "Given a matrix and a target, return the number of non-empty submatrices that sum to target." (Solution)
- "Given a rows x cols binary matrix filled with 0's and 1's, find the largest rectangle containing only 1's and return its area." (Solution)
- "Your car starts at position 0 and speed +1 on an infinite number line. (Your car can go into negative positions.) Your car drives automatically according to a sequence of instructions A (accelerate) and R (reverse)...Now for some target position, say the length of the shortest sequence of instructions to get there." (Solution)
- Recursion (12%)
- "A strobogrammatic number is a number that looks the same when rotated 180 degrees (looked at upside down). Find all strobogrammatic numbers that are of length = n." (Solution)
- "Given a binary tree, find the length of the longest path where each node in the path has the same value. This path may or may not pass through the root. The length of path between two nodes is represented by the number of edges between them." (Solution)
- Geometry / Math (11% of questions, least frequent)
- "A group of two or more people wants to meet and minimize the total travel distance. You are given a 2D grid of values 0 or 1, where each 1 marks the home of someone in the group. The distance is calculated using Manhattan Distance, where distance(p1, p2) = |p2.x - p1.x| + |p2.y - p1.y|." (Solution)
- "You are given two non-empty linked lists representing two non-negative integers. The digits are stored in reverse order and each of their nodes contain a single digit. Add the two numbers and return it as a linked list." (Solution)
Google’s line of products must be built to support 1bn+ monthly active users. Google engineers therefore need to be able to design systems that are highly scalable and performant.
While the coding questions we've covered above usually have a single optimal solution, system design questions are typically more open-ended and feel more like a conversation. Here, you’ll want to show that you can both be creative and structured at the same time.
As with coding interviews, the system design questions you'd face as a machine learning engineer should be similar to those asked in normal Google software engineer interviews. As a result, we'd recommend using the SWE questions below to prepare. For more practice, take a look at our list of 31 system design interview questions.
Example Google machine learning engineer interview questions: System design
- How would you design Google Docs?
- How would you design Google Home (voice assistant)?
- How would you design Amazon's books preview?
- How would you design a social network?
- How would you design a task scheduling system?
- How would you design a ticketing platform?
- How would you design a system that counts the number of clicks on YouTube videos?
- How would you design a webpage that can show the status of 10M+ users including: name, photo, badge and points?
- How would you design a function that schedules jobs on a rack of machines knowing that each job requires a certain amount of CPU & RAM, and each machine has different amounts of CPU & RAM? Multiple jobs can be scheduled on the same machine as long as it can support it.
So far, the interviews we've covered (coding and system design) have overlapped with the general Google software engineer interview process. The machine learning design interview, on the other hand, is specific to machine learning candidates.
Google has huge data sets and billions of users across their various apps. The magnitude and complexity of these systems present quite a few opportunities to apply machine learning to real-world problems. Broadly speaking, that's what you'll be asked to do in your interview: transform raw data, select the right model for the problem, monitor performance, etc. There may be some coding involved during machine learning rounds as well (e.g. “write code for an email auto-response system”).
The questions you'll be asked are similar to system design questions, in that you'll need to outline a high-level approach for a system or problem. The primary difference is that you'll be expected to specifically develop a machine learning solution.
To provide a clearer idea of what to expect, we've compiled a list of examples that are similar to the questions you'd be asked in Google's machine learning design interview: both the initial question and possible follow-ups as you build out the model. These questions are from three sources: Glassdoor, Rooftop Slushie, and Blind. Note that we've modified the phrasing in some cases to make the questions more clear.
Example Google machine learning engineer interview questions: Machine learning design
- How would you build, train, and deploy a system that detects if multimedia and/or ad content violates terms or contains offensive materials?
- Design autocomplete and/or spell check on a mobile device
- Design autocomplete and/or automatic responses for email
- Design the YouTube recommendation system
- How would you optimize prediction throughput for a RNN based model?
- What loss function will you optimize and why?
- What data will you collect to train your model and why?
- How will you avoid bias and feedback loops?
- How will you handle a corrupt model or an incorrect training batch?
For the behavioral interview, and sometimes at the beginning of your technical interviews, you'll be asked behavioral or "resume" questions. These questions focus on your past work experience, your qualifications, and your motivation for applying to Google. In other words, it's a way for your interviewer to get to know you better.
Behavioral questions are a great opportunity to tell your story, to highlight your top qualifications, and to demonstrate your “Googleyness.” You should also be ready to drill-down into the technical details of the projects on your resume, and to discuss the types of projects you'd like to work on in the future.
In addition, if you're interviewing for management or senior positions, you'll also usually have leadership interviews where you'll be assessed on your people and project management skills. People management questions tend to dive into how you would support and grow your team, while project management questions tend to dive into how you would effectively lead projects end-to-end.
Below, we've listed several example behavioral questions that were asked in either Google machine learning SWE or Google software engineer interviews, according to data from Glassdoor.
Example Google machine learning engineer interview questions: Behavioral
- Why Google?
- Tell me about yourself / your work at your current company
- Tell me about your machine learning background
- What are your expectations from Google?
- Tell me about a recent / interesting project you worked on
- Tell me about a time you had to resolve a conflict in a team
- What is your favorite Google product?
For management and leadership positions
- Tell me about a time you demonstrated leadership even though you weren't the formal manager
- How would you balance flexibility and process in an agile environment?
- Tell me about a time you lead a team through a difficult situation
- Tell me about a time you developed and retained team members
- How would you deal with a team challenge in a balanced way?
- How would you address a skill gap or personality conflict?
- How would you ensure your team is diverse and inclusive?
- How would you handle projects without defined end dates?
- How would you prioritize projects of varying complexity?
Now that you know what questions to expect, let's focus on how to prepare. After all, the right preparation will make the difference between failing your Google interviews and getting an offer.
Here are the four most important things you can do to prepare for Google's machine learning engineer interviews.
3.1 Learn about Google's culture
Most candidates fail to do this. But before investing tens of hours preparing for an interview at Google, you should take some time to make sure it's actually the right company for you.
Google is prestigious and it's therefore tempting to assume that you should apply, without considering things more carefully. But, it's important to remember that the prestige of a job (by itself) won't make you happy in your day-to-day work. It's the type of work and the people you work with that will.
If you know engineers who work at Google or used to work there, it's a good idea to talk to them to understand what the culture is like. In addition, we would recommend reading the following resources:
- Google's mission statement (by Google)
- Google's values (by Google)
- Google strategy teardown (by CBS Insights)
3.2 Practice by yourself
As mentioned above, you'll have four main types of interviews at Google: coding, system design, machine learning design, and behavioral. Here are resources that will help you with each question type.
For coding interviews, we recommend getting used to the step-by-step approach described by Google in the video below.
Here's a summary of the approach:
- Ask clarification questions to make sure you understand the problem correctly
- Discuss any assumptions you're planning to make to solve the problem
- Analyse various solutions and tradeoffs before starting to code
- Plan and implement your solution
- Test your solution, including corner and edge cases
To practice solving questions we recommend using our articles, 73 data structure questions and 71 algorithms questions, which have links to high quality answers to each problem. They are organized by type of data structure or algorithm as well as by difficulty level. Google has high standards for its engineers, so practice questions at medium or hard-level difficulty.
For system design interviews, we recommend studying our system design interview guide and learning how to answer system design interview questions. These guides cover a step-by-step method for answering system design questions, and they provide several example questions with solutions.
For machine learning design interviews, we recommend that you read this article from the team at Google Design on human-centered machine learning. For an end-to-end process to implement machine learning solutions, take a look at this field guide. Although the guide is from Facebook, its process breakdown should still be helpful when practicing the example questions we've provided in section 2.3 above.
For behavioral interviews, we recommend learning our step-by-step method for answering behavioral questions. Then, you can practice answering the questions listed in section 2.4 above. If you have more time to prepare, then you can prepare even more "stories" summarizing your top qualifications or important lessons that you've learned.
Finally, a great way to practice all of these interviews is to interview yourself out loud. This may sound strange, but it will significantly improve the way you communicate your answers during an interview. Play the role of both the candidate and the interviewer, asking questions and answering them, just like two people would in an interview. Trust us, it really helps!
3.3 Practice with peers
Practicing by yourself will only take you so far. One of the main challenges of machine learning interviews is communicating what you are doing as you are doing it. As a result, we strongly recommend practicing live interviews with a peer interviewing you.
If possible, a great place to start is to practice with friends. This can be especially helpful if your friend has experience with machine learning engineer interviews, or is at least familiar with the process.
3.4 Practice with ex-interviewers
Practicing with peers can be a great help, and it's usually free. But at some point, you'll start noticing that the feedback you are getting from peers isn't helping you that much anymore. Once you reach that stage, we recommend practicing with ex-interviewers from top tech companies.
If you know a machine learning engineer or someone who has experience running interviews at Google or another big tech company, then that's fantastic. But for most of us, it's tough to find the right connections to make this happen. And it might also be difficult to practice multiple hours with that person unless you know them really well.
Here's the good news. We've already made the connections for you. We’ve created a coaching service where you can practice 1-on-1 with ex-interviewers from leading tech companies. Learn more and start scheduling sessions today.