Company interview guide

General Motors Machine Learning Interview Guide (2026)

Short, behavior-heavy loop: an automated screen, practical Python and ML fundamentals, and STAR questions on GM's values.

General Motors's machine learning interview loop has 7 reported stages, from online coding assessment to behavioral and hiring-manager interviews; typical length: 2-6 weeks.

Roles covered: Data Scientist, ML Engineer and Deep Learning Engineer. The question bank holds 4 questions reported from General Motors interviews, 1 of them free to solve in your browser.

Try a free question

Interview process length
2-6 weeks
Stages
7
Reports linked to bank questions
4
Bank report dates
Nov 2022–Mar 2026
Questions in bank
4 (1 free)
Last reviewed
October 2026

How the General Motors interview loop works

Apply, then an automated online assessment (since 2025 usually a timed coding test; in 2020-2022 mostly a recorded video interview), then a 10-30 minute recruiter or hiring-manager call. The final stage is typically two interviews of about an hour, often with two interviewers each: one technical (ML and statistics questions, a resume project deep dive and, in most 2024-2026 reports, live Python coding) and one behavioral with managers. Some teams split this into more rounds or a five-person panel; one role added a case study. The company says most hires take a couple of weeks from application to offer, but several candidates waited three to six weeks for a decision after the final round. Our picture of the 2024-2026 loop rests on about a dozen reports; most of the 32 we hold are from 2015-2023.

The General Motors interview process, stage by stage

01Online coding assessment

Timed online test, about 50 min in one report, in a plain scripting environment; usually before any human contact

Recent data science candidates got an automated coding test soon after applying. Reported tasks are practical: pandas filtering and a plot written as a script rather than in a notebook, or a linear regression fitted from scratch with no scikit-learn. One candidate felt it was mainly a filter step.

What it tests

  • Basic pandas: filtering, grouping and summarizing a small dataset
  • Writing analysis code (including a plot) as a script, without a notebook
  • Implementing linear regression from first principles: normal equation or gradient descent
  • Working quickly under a fixed timer

How to prepare

  • Be able to write least squares and gradient descent for linear regression from memory in numpy.
  • Practise pandas and matplotlib in a plain .py file so the scripting setup does not slow you down.
  • Check your outputs with a quick print or assert before submitting; there is no one to ask.

Free reported questions for this stage

Automated coding test in 3 data science reports from 2025-2026 (pandas, regression from scratch). The company's hiring page says experienced applicants may be asked to complete an assessment.

02Recruiter or hiring-manager call

10-30 min phone or video call with a recruiter; some teams have a short hiring-manager call first

A screen on your background, why you want to work at GM, location and salary. Some candidates spoke to the hiring manager first, with no technical questions. One candidate warns that the salary figure you agree to here is held to in the offer, and several who needed visa sponsorship say it ended their process, so raise it early.

What it tests

  • How your experience matches the posting
  • Why GM, and why this team
  • Logistics: location, salary expectations, work authorization

How to prepare

  • Give a considered salary range; candidates say the number you accept here is used in the offer negotiation.
  • Confirm the location of the role in writing; one candidate was later offered a different site.
  • Ask what each later round covers; one recent candidate noted clear written instructions from the recruiter.

Recruiter or hiring-manager call in 13 data science, ML and vision reports (2015-2026), 6 since 2023; length 10-30 min where stated. The hiring page names a recruiter as contact, not a call.

03Recorded video interview (older loops)

One-way recorded video: about 4-5 questions, roughly 1 min to prepare and 1-3 min to answer, usually one retake

Common in 2020-2022 data science and ML loops, sometimes sent right after applying and sometimes after the recruiter call. Most questions are tell-me-about-a-time behavioral prompts; a few senior data science screens added short technical questions. One version included logic games. No report since 2023 mentions it, so treat it as possible rather than expected.

What it tests

  • Teamwork and conflict stories in STAR form
  • Why your background fits the role
  • Short spoken answers on overfitting, correlation versus causation and hypothesis testing

How to prepare

  • Prepare five or six STAR stories and rehearse them on camera with a timer.
  • Use the retake if your first answer runs over; only the final recording is kept.
  • Look for the accommodation option before you start; one candidate found it only after clicking through.

Recorded video screen in 11 data science and ML reports, all from 2020-2022; none of the 2023-2026 reports mention it. Not described on the company's hiring page.

04Live Python coding

Part of the technical interview or one or two separate rounds, in a plain editor or a notebook

Most 2023-2026 reports include some live coding, and the level is usually basic: loops and dictionaries, counting words and their positions in a string, array questions, printing even numbers. Vision and research roles add numpy and matplotlib, and one computer vision team asked for vectorized numpy with no explicit loops. A recent senior data science loop had several long coding rounds.

What it tests

  • Core Python: loops, dictionaries, string splitting and counting
  • Array manipulation without an IDE
  • Vectorized numpy for image or geometry problems
  • Quick data analysis and plotting on a sample dataset

How to prepare

  • Practise easy string and array problems in a plain text editor with no autocomplete.
  • For vision roles, rewrite loop-based numpy code with broadcasting and explain the speed-up.
  • Talk through your approach; one panel said it cared mostly about the approach.

Live coding in 7 data science, deep learning, vision and research engineer reports from 2023-2026; difficulty described as basic in most, longer rounds in one 2026 senior loop.

05ML and statistics technical interview

About 60 min with one or two data scientists or managers; a five-person panel in one research engineer loop

Conversational questions on fundamentals, often tied to how you would solve a problem. Data science teams ask about regression metrics, overfitting and model comparisons; deep learning and vision teams ask about loss functions, activation functions, rotation matrices and projection; one research engineer panel also asked about Git, CI/CD and Docker. Interviewers tend to guide you through the problem and probe with follow-ups.

What it tests

  • Regression basics: R-squared, assumptions, overfitting
  • Comparing models such as SVM and random forest, and how you tuned them
  • How you would frame and approach a new ML problem
  • Deep learning and vision basics: L1 vs L2 loss, activations, rotation matrices, forward projection
  • Engineering hygiene for research roles: version control, CI/CD, containers

How to prepare

  • Revise classical ML and regression thoroughly; it comes up more than deep learning.
  • For every model on your resume, be ready to say why you chose it over the obvious alternatives.
  • For vision roles, review rotation matrices, quaternions versus Euler angles and the camera projection model.

ML or statistics technical questions in 9 data science, deep learning, vision and research reports (2020-2025); topics vary by team and most describe the level as moderate.

06Project deep dive

Usually inside the technical or behavioral hour; sometimes its own interview with two or three people

A detailed walk through one or two projects from your resume. Interviewers let your answers steer the follow-ups: why this method, why not the alternatives, which architectures and losses you tried. Several candidates said this took most of the interview.

What it tests

  • Explaining a project end to end and your own part in it
  • Justifying method, architecture and loss choices against alternatives
  • Offering a different approach to the same problem
  • How you get models accepted and used by stakeholders

How to prepare

  • Pick two projects and prepare the rejected alternatives and why you dropped them.
  • Only list techniques you can defend in detail; the follow-ups go wherever you lead.
  • Have one example of a model you took to production or to a business decision.

Resume project deep dive in 8 data science, deep learning, vision and AI scientist reports (2016-2025); often combined with behavioral or technical questions in the same hour.

07Behavioral and hiring-manager interviews

One or two interviews of about 60 min, often two managers at once; one hiring-manager interview ran 90 min after a case study

The most consistent part of the loop, and in some ML loops the only live interviews. Expect competency questions in STAR form on conflict, failure, risk and integrity, plus why GM. The company says its interviews test its values and behaviors, so expect questions on speaking up and treating people fairly.

What it tests

  • Conflict with a coworker and challenging projects
  • Failure, calculated risk and what you learned
  • Integrity: what you would do if a colleague were treated unfairly
  • Why GM, and fit with the team you would support

How to prepare

  • Read GM's published values and behaviors and map one story to each.
  • Answer in STAR form with real, job-related examples, as the company's hiring guide asks.
  • Prepare an answer for seeing someone mistreated when the manager does not step in; two candidates got it.
  • Follow up with the recruiter after the final round; several candidates waited weeks for an answer.

Live behavioral or leadership interview in 12 data science, ML and AI scientist reports (2016-2025). The company's hiring page says interviews are competency-based and asks for STAR answers.

Questions and themes reported most often

  • Tell me about a time: conflict, failure, a challenging project or a calculated risk (7 reports, behavioral and hiring-manager interviews)
  • Why GM, and why this position (3 reports, recruiter or hiring-manager call)
  • What you would do if someone were treated unfairly and the manager did not step in (2 reports, behavioral and hiring-manager interviews)
  • Fit linear regression from scratch without scikit-learn (1 reports, online coding assessment)
  • Pandas filtering and a plot in a scripting environment (1 reports, online coding assessment)
  • SVM versus random forest, and how you set the forest's parameters (1 reports, ml and statistics technical interview)
  • What R-squared tells you about a regression (1 reports, ml and statistics technical interview)
  • Overfitting, correlation versus causation and running a hypothesis test (1 reports, recorded video interview (older loops))

What General Motors asks about

  • Behavioral & GM values30%
  • ML & statistics fundamentals25%
  • Python & data coding22%
  • Project deep dive15%
  • Computer vision & linear algebra8%

General Motors interview prep checklist

  • Wrote linear regression from scratch in numpy in under 20 minutes
  • Solved a pandas filter-and-plot task in a plain .py file
  • Prepared six STAR stories covering conflict, failure, risk, integrity and a hard project
  • Can explain R-squared, overfitting, and SVM versus random forest without notes
  • Walked through two resume projects with the alternatives you rejected
  • Have a specific answer to why GM and this team
  • Agreed salary range, location and work authorization with the recruiter

Tips

  • Spend at least as much time on STAR stories as on technical prep; behavioral rounds appear in almost every report.
  • Map your stories to GM's published values and behaviors before the final interviews.
  • Drill linear regression and pandas in a plain script; the recent online test uses that setup.
  • Revise classical ML and regression metrics before deep learning, unless the role is in vision or autonomy.
  • For vision roles, practise vectorized numpy and the geometry of rotations and projection.
  • Prepare two resume projects in depth, including the alternatives you rejected.
  • Settle salary, location and sponsorship questions clearly with the recruiter early on.
  • Expect slow feedback after the final round and follow up politely with your recruiter.

See the rest of the General Motors loop

The later stages, the questions reported for them and how to prepare are in Pro.

Unlock with Pro

Practice the General Motors loop

Work the stages in order: every reported question for this company, grouped by round, with solutions in the app.

General Motors interview FAQ

How many interview rounds does General Motors have for ML roles?

Candidates report 7 stages: Online coding assessment, Recruiter or hiring-manager call, Recorded video interview (older loops), Live Python coding, ML and statistics technical interview, Project deep dive and Behavioral and hiring-manager interviews. Apply, then an automated online assessment (since 2025 usually a timed coding test; in 2020-2022 mostly a recorded video interview), then a 10-30 minute recruiter or hiring-manager call. The final stage is typically two interviews of about an hour, often with two interviewers each: one technical (ML and statistics questions, a resume project deep dive and, in most 2024-2026 reports, live Python coding) and one behavioral with managers. Some teams split this into more rounds or a five-person panel; one role added a case study. The company says most hires take a couple of weeks from application to offer, but several candidates waited three to six weeks for a decision after the final round. Our picture of the 2024-2026 loop rests on about a dozen reports; most of the 32 we hold are from 2015-2023.

How long does the General Motors ML interview process take?

Reported timelines run 2-6 weeks.

What does General Motors ask in ML interviews?

By reported question mix: Behavioral & GM values 30%; ML & statistics fundamentals 25%; Python & data coding 22%; Project deep dive 15%; Computer vision & linear algebra 8%. The most repeated themes: Tell me about a time: conflict, failure, a challenging project or a calculated risk; Why GM, and why this position; What you would do if someone were treated unfairly and the manager did not step in.

How should I prepare for the General Motors ML interview?

Spend at least as much time on STAR stories as on technical prep; behavioral rounds appear in almost every report. Map your stories to GM's published values and behaviors before the final interviews. Drill linear regression and pandas in a plain script; the recent online test uses that setup.

Are there free General Motors interview questions?

Yes. LeetML has 4 questions reported from General Motors interviews, and 1 of them are free to solve in the browser; they are listed by stage above.

Related interview guides

Official sources