Company interview guide
Workday Machine Learning Interview Guide (2026)
Recruiter and hiring-manager calls, a practical Python or pandas screen, then coding, ML theory and design rounds.
Workday's machine learning interview loop has 7 reported stages, from recruiter or hr call to behavioral and project discussion; typical length: About 4-6 weeks, sometimes much longer.
Roles covered: ML Engineer and Data Scientist. The question bank holds 1 question reported from Workday interviews.
- Interview process length
- About 4-6 weeks, sometimes much longer
- Stages
- 7
- Reports linked to bank questions
- 1
- Bank report dates
- Sep 2015–Sep 2015
- Questions in bank
- 1 (0 free)
- Last reviewed
- October 2026
How the Workday interview loop works
Most reports open with a short recruiter or HR call and a conversation with the hiring manager, then a technical screen. In 2025 ML-engineer reports the screen covers resume and ML basics or Python and pandas data manipulation, and a full loop can add separate rounds for ML coding, software system knowledge, pandas, ML theory and the hiring manager. Senior and older loops describe an onsite of four or five interviews, or one long technical session, with coding, design and a leadership conversation at the end. Two data science loops from before 2020 used a multi-day data-analysis take-home instead of a live screen. The company's careers page says only that there are several phone, video or in-person rounds. We hold under 20 first-hand reports and only about six since 2021, so the order and formats here are a best reading of thin evidence. Several candidates waited weeks for feedback, so follow up with your recruiter.
The Workday interview process, stage by stage
01Recruiter or HR call
A short phone or video call with a recruiter or HR partner; one candidate reports about 10 minutes
A first conversation about your background, your projects and why you are interested in Workday. Some recruiters reach out on professional networks and introduce you to the hiring manager straight after. One 2025 senior data science candidate says this round already included basic behavioral questions.
What it tests
- A clear summary of your background and recent projects
- Why Workday and what you know about its products
- Logistics: role, level, location and timeline
How to prepare
- Ask which rounds follow and whether the screen is live coding, pandas work or a take-home.
- Have a two-minute walk through your most relevant ML or data project ready.
Recruiter or HR call first in 7 ML and data science reports, four of them since 2021. The company's intern hiring post lists a recruiter or hiring-manager conversation.
02Hiring manager conversation
One call early in the process and often a second at the end; senior loops may close with an executive
Most loops include the hiring manager twice: a call before the technical rounds that looks at fit and past work, and a closing conversation after them. One 2025 candidate had an extra hiring-manager round because a stand-in took the first one, and a 2025 senior candidate finished with the CTO. One 2026 candidate describes the final hiring-manager talk as short and rushed.
What it tests
- How your past work maps to the team's product and data
- Motivation for the role and for Workday
- Ownership and depth behind your resume claims
How to prepare
- Learn what the team builds (for example HR, finance or planning products with ML features) and tie your projects to it.
- Prepare questions about the team's data, models and how they ship to customers.
- Treat the final call as a real evaluation even if it is brief.
Hiring-manager rounds in 8 ML and data science reports, four since 2025, usually early and again at the end. The company's intern hiring post names a hiring-manager or recruiter conversation.
03Technical screen: Python, pandas and ML basics
A phone or video screen with live coding, about 45 min in one report; pandas tasks may run against hidden test cases.
Two older data science loops used a multi-day data-analysis take-home instead
Recent screens focus on practical data work rather than puzzles: parse records in Python, filter and clean a DataFrame, apply functions row by row and manipulate strings. One 2025 data science candidate passed most but not all hidden tests and was not moved forward, so completeness matters. Other screens open with your resume and ML basics. Older screens mixed a small coding problem with data science and statistics questions.
What it tests
- Pandas filtering, dropping rows, lambdas and string handling
- Parsing and preprocessing raw data in plain Python
- ML basics tied to your resume, including any LLM experience
- Writing correct code without seeing the test cases
How to prepare
- Drill everyday pandas until it is automatic: boolean masks, apply, groupby, str methods and missing values.
- Ask how the code will be checked; if tests are hidden, write your own edge-case checks before submitting.
- Be ready to say what you have built with LLMs; it came up in a 2025 screen.
Phone or video technical screen in 9 ML and data science reports; the four since 2021 centre on Python or pandas data work. Two pre-2020 loops used a take-home. The company's intern post describes a technical interview.
04Coding and data structures
Whiteboard or online coding in a screen or onsite round; reported difficulty runs from easy to hard
Classic coding questions appear across many reports, though most of them date from 2015 to 2019. Merging sorted lists or arrays recurs, with follow-ups to do it in place or bring the cost from quadratic to linear. Data science candidates have been surprised by how algorithm-heavy these rounds were, and one onsite ended early after two coding rounds went badly.
What it tests
- Array and list manipulation such as merging sorted inputs
- Improving the time or space complexity of a working solution
- Hash-based lookups such as finding pairs that sum to a target
- Clean, correct code written by hand
How to prepare
- Practise easy and medium array and two-pointer problems, then their in-place variants.
- State the complexity of your first solution and offer the faster one before being asked.
Data-structure coding questions in 8 ML and data science reports, all but one from 2015 to 2021; merging sorted lists or arrays appears in three of them.
05ML coding, theory and statistics
Separate ML coding and ML theory rounds in a 2025 loop; older loops asked ML and probability questions inside technical rounds
The fullest 2025 ML-engineer report lists an ML coding round and an ML theory round that discussed theory alongside real use cases. Older data science reports asked about hypothesis tests, p-values, choosing between classifiers such as logistic regression and a neural network, and short probability questions that tripped up some candidates.
What it tests
- Core ML theory and when to use which model
- Applying ML to a concrete business use case
- Hypothesis testing and p-values
- Quick probability reasoning, explained step by step
How to prepare
- Review classic models, their assumptions and failure modes, and practise comparing two of them aloud.
- Work short probability questions on paper and narrate the process, not only the answer.
ML, statistics or probability questions in 7 ML and data science reports; only two (2021 and 2025) are recent, and the 2025 one names separate ML coding and theory rounds.
06Software and system design
A software system knowledge round, a design round with a product manager, or an object-oriented class-diagram exercise
ML-engineer loops include some engineering design, but formats differ: a 2025 loop had a software system knowledge round, a 2021 onsite had a system design round with a product manager, and a 2022 loop asked for a class diagram of an email system. Expect to show that you can build production software, not only models.
What it tests
- Object-oriented modelling of a small system
- How the parts of a production service fit together
- Explaining design choices to a non-engineer
How to prepare
- Practise drawing a class diagram for a messaging or similar system with clear responsibilities.
- Be ready to discuss how a model is served and maintained inside a larger application.
Design rounds in 3 ML-engineer reports from 2021, 2022 and 2025, each in a different format.
07Behavioral and project discussion
Standalone behavioral interviews or questions folded into other rounds; one loop ended with a demonstration
Expect to talk about your favourite project, the data analysis you have done and why you want to work at Workday. A 2025 intern loop had a behavioral stage before the technical one, and a 2022 ML-engineer loop had two behavioral interviews plus a final demonstration.
What it tests
- Depth on one or two past projects
- Why Workday
- Collaboration and debugging stories with clear outcomes
How to prepare
- Use situation, task, action and result to structure stories; the company's own hiring post recommends it.
- Prepare one project in detail: data, model choice, results and what you would change.
Behavioral or project questions in 6 ML and data science reports, three since 2022. The company's intern hiring post describes behavioral questions answered in a structured way.
Questions and themes reported most often
- Clean and filter data in Python or pandas: drop rows, apply functions, handle strings (4 reports, technical screen: python, pandas and ml basics)
- Merge two sorted lists or arrays, then in place or in linear time (3 reports, coding and data structures)
- Why Workday and what you know about the company (2 reports, behavioral and project discussion)
- Short probability questions inside a technical round (2 reports, ml coding, theory and statistics)
- Data-analysis take-home on a provided dataset (2 reports, technical screen: python, pandas and ml basics)
- Hypothesis testing, p-values and logistic regression versus a neural network (1 reports, ml coding, theory and statistics)
- Class diagram for an email system that sends and receives messages (1 reports, software and system design)
- Find pairs in an integer array that sum to a target (1 reports, coding and data structures)
What Workday asks about
- Coding & data structures30%
- Python & pandas data work25%
- ML & stats concepts20%
- Behavioral & hiring manager15%
- Software & system design10%
Workday interview prep checklist
- Cleaned and filtered a messy DataFrame with masks, apply and string methods without looking anything up
- Merged two sorted arrays in place and explained the complexity
- Can compare logistic regression with a neural network and explain a p-value plainly
- Drew a class diagram for a small messaging system
- Have a two-minute project summary and a specific answer to why Workday
- Asked the recruiter which rounds and formats to expect
Tips
- Ask the recruiter early what each round covers; formats vary by team and our evidence is thin.
- Make everyday pandas fast and exact; a recent screen was graded on hidden test cases.
- Practise easy and medium array problems with in-place and linear-time follow-ups.
- Review ML theory with use cases, plus hypothesis tests and basic probability.
- Prepare one engineering design exercise, such as a class diagram for a messaging system.
- Research the team's product area and connect your projects to it for both hiring-manager calls.
- Have a ready answer about any work you have done with LLMs.
- Follow up politely if feedback is slow; several candidates waited weeks.
Practice the Workday loop
Work the stages in order: every reported question for this company, grouped by round, with solutions in the app.
Workday interview FAQ
How many interview rounds does Workday have for ML roles?
Candidates report 7 stages: Recruiter or HR call, Hiring manager conversation, Technical screen: Python, pandas and ML basics, Coding and data structures, ML coding, theory and statistics, Software and system design and Behavioral and project discussion. Most reports open with a short recruiter or HR call and a conversation with the hiring manager, then a technical screen. In 2025 ML-engineer reports the screen covers resume and ML basics or Python and pandas data manipulation, and a full loop can add separate rounds for ML coding, software system knowledge, pandas, ML theory and the hiring manager. Senior and older loops describe an onsite of four or five interviews, or one long technical session, with coding, design and a leadership conversation at the end. Two data science loops from before 2020 used a multi-day data-analysis take-home instead of a live screen. The company's careers page says only that there are several phone, video or in-person rounds. We hold under 20 first-hand reports and only about six since 2021, so the order and formats here are a best reading of thin evidence. Several candidates waited weeks for feedback, so follow up with your recruiter.
How long does the Workday ML interview process take?
Reported timelines run About 4-6 weeks, sometimes much longer.
What does Workday ask in ML interviews?
By reported question mix: Coding & data structures 30%; Python & pandas data work 25%; ML & stats concepts 20%; Behavioral & hiring manager 15%; Software & system design 10%. The most repeated themes: Clean and filter data in Python or pandas: drop rows, apply functions, handle strings; Merge two sorted lists or arrays, then in place or in linear time; Why Workday and what you know about the company.
How should I prepare for the Workday ML interview?
Ask the recruiter early what each round covers; formats vary by team and our evidence is thin. Make everyday pandas fast and exact; a recent screen was graded on hidden test cases. Practise easy and medium array problems with in-place and linear-time follow-ups.
Are there free Workday interview questions?
LeetML has 1 question reported from Workday interviews. The stage sections above link the free practice problems for the topics each round covers; the reported-question set itself is in the app.
Related interview guides
- SalesforceA team-run loop that mixes plain coding, ML and LLM fundamentals, applied design and a deep look at your own work.
- OracleRecruiter and hiring-manager screens, then a technical screen and a multi-round loop mixing practical coding, ML and LLM fundamentals, system design and behavioral questions.
- SAPAn online coding test, an HR screen, then project-driven technical rounds with easy-to-medium coding and a hiring-manager conversation.
- IntuitA data-science-heavy loop built around the craft demo: model a dataset, then defend it to a panel.
- AdobeMedium-level algorithm coding, practical ML fundamentals and deep project discussion, with a job talk for research roles.
- LinkedInProbability-flavoured coding, ML fundamentals and recommender design built around LinkedIn's own products.