Intuit Data Scientist Interview Guide

Intuit Inc. is one of the biggest small business and financial technology companies in the world. The company develops and sells business and financial management software solutions (QuickBooks), tax solutions for individuals (TurboTax), and personal finance solutions (Mint and Credit Karma now). Founded in 1983, Intuit has since emerged as a leading fin-tech company with over 50 million customers served worldwide in over nine countries.

Intuit generates tons of customer data yearly, connecting all of its products together. As a data-driven company, data science is at the core of everything, and Intuit has over the years been leveraging data science in advanced analytics and machine learning tools to improve their customers’ financial lives.

Data scientist roles at Intuit vary across different teams, and the specific roles of a data scientist within each team will be heavily determined by the needs of that group. From teams such as Small businesses to Machine Learning Futures, data scientist teams at Intuit analyze data and deploy ML and AI models to solve business-related problems. Generally speaking, the scope of data science at Intuits spans from business analytics and data engineering, and the tools used may range from basic analytics to machine learning and deep learning.

Required Skills

Intuit’s preferred data science hiring requirements may vary across specific teams and groups, but in general, hire only talented and qualified applicants with a minimum of 3 years (5+ years for senior-level) in data science roles.

Other basic requirements for hiring include:

• BS, MS, or PhD in Statistics, Applied Math, Operations Research, Computer Science, Physics, Engineering, or related fields, or equivalent experience.
• Knowledgeable with data science tools and frameworks (i.e., Python, Scikit, NLTK, Numpy, Pandas, TensorFlow, Keras, R, Spark).
• 1 to 3+ (5+ years for senior-level) years’ experience with a general-purpose programming language (e.g., Python, C, Java, etc.).
• Domain expertise in subjects such as experimental design and multivariate ab testing.
• Strong interpersonal and communication skills in order to effectively contribute to technical teams and make presentations to a variety of technical and business personnel.

What kind of data science role?

Data science roles at Intuits are spread across a wide range of groups. On the surface, a data scientist at Intuit is someone who uses advanced analytics tools, machine learning, NLP, and AI algorithms to provide business-impact recommendations. However, specific roles may span from product-specific analytics teams embedded on a team to machine learning engineering implementation. Depending on the group assigned, the functions of a data scientist or machine learning engineer at Intuit may include:

• Risk Research & Intelligence: Building and prototyping algorithms and applications to improve security and anti-fraud on top of the collective financial data of 60 million consumers and small businesses
• Smart Money Services: Leveraging data mining and machine learning techniques to manage credit and fraud risk in payments and payroll.
• Small Business Data Science team: Using industry-leading analytics tools and techniques to drive user growth and retention in small businesses.
• Core Data Science Team: Develop, design, and integrate ML models into production. Collaborate and build AI solutions for all internal teams e.g. Engineering, HR, Finance & Legal, etc.
• Customer Success Data Team: Pull out insights from customer success data and apply them to all intuits products (TurboTax, QuickBooks, Mint, etc.).

Intuit’s data science interviews start with an initial phone call from a recruiter, followed by a technical video interview of past relevant projects and a take-home challenge. After finishing through the initial stages, an onsite interview will be scheduled, which consists of four 45-minute long interviews with various team members, technical manager, and the product manager.

Initial Screen

The initial interview is a resume-based phone interview with an HR or recruiter. This interview aims to assess your skills and past projects to see if you are a great fit for the team you are applying for. Questions in this screening are standard resume-based questions.

Technical Screen

The technical screen at Intuit is after the recruiter screen. It is done either with Karat, an external interviewing service, or with an Intuit hiring manager. Interview questions for data science roles consists of testing analytics and coding skills in SQL and Python, respectively.

Here’s a sample question that you can try:

Let’s say 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 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.

How would we build this model?

The interview is an hour-long, and it’s pertinent to display a clear aptitude for technical ability. The interviewer will also go over any past projects to get a sense of your past experience. Really nail down your resume and how to talk about your projects in-depth and how they relate to applied machine learning.

The Take-Home Challenge

Intuit gives a data challenge before the onsite interview, and applicants are required to complete this within four hours of receiving the take-home. The take-home challenge comprises a standard Intuit case study dataset on TurboTax. You’ll have to run analytics in SQL and work on a machine learning problem on the dataset.

The Onsite Interview

The onsite interview at Intuit comprises four interview rounds (two technical, one data-challenge presentation, and one behavioral). Technical questions in this interview are mainly open-ended and span across basic statistical concepts (A/B Testing), modeling, experimental design, SQL, and machine learning algorithms. In general, the onsite interview at Intuit looks like this:

• Data challenge presentation: In this round, candidates are expected to create dashboards of the data given and coherently explain to the interviewer the type of analysis done and how well the data was explored. This will require coding in a live environment and analyzing the data while talking through what you’re doing.
• Technical interview with a technical manager talking about experience and machine learning concepts.
• Technical interview with a data scientist that involves coding SQL and algorithms plus probability and statistics questions.
• Hiring manager interview This round is your standard behavioral interview. It’s essential to be well-prepared, and, ideally, use the STAR framework when answering behavioral questions.

At Interview Query, we love to hear from those who’ve successfully landed jobs in the data science field. To help the rest of our community, we’re sharing their career path stories and approaches to interview preparation.

We caught up with Owen McCarthy, who joined Intuit after completing his Bachelor’s in Data Science at UCSD in 2020 and followed an unconventional path! We discussed his personal journey, tips for getting to the interview stage of applications, and the Intuit interview guide.

What was your journey into data science?

I attended the University of California, San Diego, which didn’t have a data science major when I started. Instead, I blended the computer science and business majors to build a bridge between them and was lucky that they kicked off a dedicated DS program my sophomore year. I was part of the inaugural cohort to graduate from this new program and also snagged a business minor on the way out.

I started off looking for roles in data science, with a strong focus on natural language processing. OpenAI and GPT-2 were just coming out, and I started off looking for roles in data science, with a strong focus on natural language processing. OpenAI and GPT-2 were just coming out, and by the time GPT-3 launched, I knew that this was a field worth putting a real bet on career-wise. The roles that companies were hiring for wanted more extensive educational backgrounds than mine, but I got lucky with a data science program with a company called Barisk.

The program was designed as a rotational, where every 18 months, you would be moved to a new business sector and geographic area. After three rotations, or three and a half years, you come out as a senior data scientist, a project manager, or a data science manager.

Even though it was still geared towards master’s graduates, I noticed a small input at the bottom of the application, which allowed you to communicate extra information to the hiring managers. In this way, I was able to overcome the lack of a higher degree by speaking with the team more directly about my interest in the field.

How did you land an interview with Intuit?

I did not go through the regular channels to get the interview. I paid for LinkedIn Premium, searched for data science recruiters, and emailed them directly if they had an email in their bio. You can also try to hunt their email down on the web if you know their name and company. There are quite a few websites for that.

I would email these individuals, letting them know that I was interested in their group and that I had experience as a data scientist. I made sure to attach my resume as well and saw quite a bit of success with this method.

Intuit eventually got back to me on a senior data science position, and I started preparing for my interview there!

How did you prepare for the interview?

Preparing for data science interviews can be tricky since there’s so much breadth of content. You’re being tested on Python or R knowledge, SQL, small data structures, stats and probability, machine learning, and some business or product questions. There is just so much out there to know.

For the current role, I studied SQL for my interviews, since that’s what they advertised they were looking for in the job listing. I also reviewed quite a bit of the theory behind general machine learning algorithms. Some examples are:

• How does a neural network generally work?
• What’s the difference between different neural network architectures (RNN, CNN, Transformer, etc.)?
• How does SVM (Support Vector Machines) work?
• What’s a linear regression model?
• What’s the difference between boosting and bagging?

Knowing these foundational machine learning concepts proved to be really helpful.

Lastly, my big tip would be while it’s essential to practice Python and SQL, don’t overlook the importance of preparing for behavioral interviews. What I observed with my data science colleagues is that they typically ace the technical portions, but will get leveled if their behavioral answers aren’t strong. Always use the STAR method (situation, task, action, and result), be thoughtful, and answer the prompt.

What was your experience interviewing at Intuit?

Recruiter Call (30 Minutes)

This was your regular thirty-minute call, to determine what roles I was qualified for and if I could work from the Mountain View campus or remotely. We also discussed salary and benefits for certain roles, which I stayed non-committal on, pending the final position and scope.

Technical Screening (1 Hour)

I had one Python question and two SQL questions, which took around 30 to 40 minutes to complete.

There was then around 20 minutes to discuss my background.

If there is extra time, they’ll likely ask some filler questions on Python or general machine learning, basic topics like bias, variance, trade-off, boosting, or bagging.

The Four Round Interviews (3-4 Days)

Round 1: Solution Creation and Demonstration

You are given a problem, and you need to create a machine-learning solution to demonstrate and present. That presentation is to the team and needs to be around an hour in length. That’s a lot of time to be speaking to this solution, so you definitely need to do quite a bit of prep for it. They also spent around ten minutes on the candidate’s background as a chance to get to know you better.

Round 2: Skip-Level Manager (The Boss’ Boss)

This interview is with your supervisor’s manager and is all about stakeholder management. They want to know how you go about product-related projects, how you work with others, your formal or informal leadership style, and what type of manager you’re looking for. There might also be a few business questions peppered in here.

Round 3: Technical Interview

Another hour-long technical assessment, again focused on conversational-style questions, Python, SQL, and your background again.

Round 4: Hiring Manager

This is just a typical behavioral interview, with nothing too out of the ordinary. Don’t forget your preparation and the simple STAR framework.

Note: The exact order of the final four interviews may be different for each candidate.

• How does boosting work?
• Let’s say you can play a coin-flipping guessing game either once or a 2 out of 3 games. What is the best strategy for winning?
• Given a long array that you can’t store, how do you find the median?
• What are the limitations of linear regression?
• Describe how a random forest works under the hood.
• Implement an iterator function that takes three iterators as the input and sorts them.
• What features would you add to a model that doesn’t already exist?
• If there was a feature that 100% of the users used, would it be a good feature?
• What’s an adequate rebalance of an imbalanced dataset?

See more Intuit data scientist questions from Interview Query