Microsoft AI/ML Intern Interviews: 114 Experiences (Part 1 of 2)
- Company
- Microsoft
- Reports
- 114 interview reports
- Level
- Internship
- Stages
- Phone screen → Onsite
Microsoft AI/ML intern interviews for data science, applied science and research roles mix coding, ML fundamentals and project deep dives.
At a glance
- Process: an initial screen, then often three back-to-back interviews (006, 008, 009, 012, 017, 018). Two of the 16 current reports had an online assessment first (008, 009, on HackerRank in 009).
- Asked recently: coding in 13 of the 16 current reports, ML and LLM fundamentals in 9, a research or project deep dive in 9, and a design question (RAG, counterfactual evaluation, LLM design) in 5. Named coding included Top K Frequent Elements (005), the median of two arrays (018), K-means from scratch (007) and a precision-recall curve (009).
- Difficulty: the coding was mostly easy (006, 016, 019) to medium (017); 018's interviewer kept asking for optimizations, and 012 found the rounds more data-science-focused than they had prepared for.
- Practise first: Implement a Precision-Recall Curve (named in three reports across the years), Implement K-Means from Scratch, Top K Frequent Elements and Find the Median from Two Arrays.
Coverage: 114 Microsoft AI/ML internship reports (reports 005–118): Data Scientist, Data & Applied Scientist, Applied Scientist, Research and Student Researcher interns, at phone or technical screens and final loops, from 2020 to 2026. The 16 reports from Nov 2023 on (005–012, 014–021) are current; the other 98 (2020–Oct 2023) are under History, outside the current counts. Report 013's interview predates Nov 2023.
Related: Reports 001–004 are in 2025–2026 Microsoft CoreAI / AI-ML SWE Intern Interviews, which focuses on the CoreAI / AI-ML Software Engineering Intern loop. This article covers other Microsoft AI/ML internship reports without treating them as CoreAI SWE evidence. Together they make up the 118 experiences counted below.
What came up
These counts cover the 16 current reports. Coding is in almost every loop, but ML knowledge and the candidate's own research carry as much weight.
| Topic | Reports | Last seen |
|---|---|---|
| Coding (LeetCode-style or from scratch) | 13 of 16 | Sep 2026 (report date) |
| ML and LLM fundamentals | 9 of 16 | Feb 2026 |
| Research or project deep dive | 9 of 16 | Feb 2026 |
| Design (ML, RAG or product) | 5 of 16 | Feb 2026 |
| Metrics, experiments and statistics | 5 of 16 | Feb 2026 |
| Behavioral questions | 4 of 16 | Nov 2025 (report date) |
| Online assessment | 2 of 16 | Feb 2026 (report date) |
Rows: coding 005–010, 012, 014, 016–019, 021; fundamentals 006, 007, 009, 010, 012, 015, 016, 018, 019; research or project 006, 007, 009, 011, 015, 016, 018, 019, 020; design 006, 008, 018, 019, 020; metrics and statistics 006, 007, 008, 009, 018; behavioral 011, 014, 017, 020; OA 008, 009.
Questions we found
These tables cover the 16 current reports. Older questions are in the History tables.
Algorithms / Data Structures
| Question | Reports | Last seen |
|---|---|---|
| Top K Frequent Elements | 1: 005 | Sep 2026 (report date) |
| Return Every Pair with the Minimum Numeric Difference | 1: 006 | Feb 2026 |
| Show an Organization Chart One Layer at a Time | 1: 008 | Feb 2026 (report date) |
| Explain How ChatGPT Is Trained | 1: 012 | Oct 2025 |
| Find a Median with a Memory-Constrained Follow-Up | 1: 017 | Nov 2024 (report date) |
| Find the Median from Two Arrays | 1: 018 | Oct 2024 |
| Expected Exit Time with Returning Paths | 1: 018 | Oct 2024 |
| Reverse Word Order In Place | 1: 021 | Jan 2024 (report date) |
ML and statistics coding
| Question | Reports | Last seen |
|---|---|---|
| Compute Dog Retrieval Precision and Recall | 1: 006 | Feb 2026 |
| Implement K-Means from Scratch | 1: 007 | Feb 2026 (report date) |
| Implement a Precision-Recall Curve | 1: 009 | Nov 2025 |
ML and system design
| Question | Reports | Last seen |
|---|---|---|
| Design a Retrieval-Augmented Generation System | 1: 006 | Feb 2026 |
| Design Counterfactual Evaluation When A/B Testing Is Unavailable | 1: 008 | Feb 2026 (report date) |
| Design How a Company Should Use AI | 1: 019 | May 2024 |
ML, data science and product cases
| Question | Reports | Last seen |
|---|---|---|
| Diagnose a Sudden Product-Metric Drop | 1: 008 | Feb 2026 (report date) |
16 candidate reports, Nov 2023 to 2026
005: Copilot Research Intern — phone or technical screen
Report date: 2026-09-27 · Role: Copilot Research Intern · Stage: Phone / technical screen
Microsoft Copilot Research Intern report describing a pure coding video interview, one exact LeetCode problem, interviewer hints, two solution families, complexity discussion, and a pass result.
- Top K Frequent Elements. The exact task is LeetCode 347, Top K Frequent Elements. A hashmap-plus-heap solution is O(n log k), while bucket sort can be O(n). The interviewer may ask complexity and optimization follow-ups. Practice this problem
006: Applied Science PhD Intern — phone or technical screen
Report date: 2026-02-23 · Role: Applied Science PhD Intern · Stage: Phone / technical screen · Interview date: 2026-02-19
Detailed Microsoft Applied Science PhD Internship Opportunities - Redmond report with one initial screen, three back-to-back interviews, project and ablation-study discussion, NLP/ML fundamentals, two coding tasks, RAG design, and an offer.
- Return Every Pair with the Minimum Numeric Difference. Given [-1,-2,-4,-5,7,10], find the minimum difference and return every pair achieving it. The reported output is [[-5,-4],[-2,-1]]. The interviewer described the task as Easy. Practice this problem
- Compute Dog Retrieval Precision and Recall. There are ten image files, each containing an animal image. An API returns k files; compute precision and recall for dog images. Handle abnormal API results such as None or a result count different from k. Practice this problem
- Design a Retrieval-Augmented Generation System. Design a RAG system. The candidate discussed an agentic-tool framing.
Other reported topics or limits: Ablation Design, RL, Agents, Metrics, and Regularization.
007: Applied Scientist Intern — final interview loop
Report date: 2026-02-13 · Role: Applied Scientist Intern · Stage: Final loop / onsite
Microsoft Applied Scientist internship report covering a phone screen and three virtual-onsite rounds with ML/LLM fundamentals, from-scratch K-means, hypothesis testing, Python debugging, project discussion, case study, and an offer.
- Implement K-Means from Scratch. Implement K-means from scratch. Discuss the algorithm's time complexity. Practice this problem
Other reported topics or limits: ML and LLM Fundamentals Phone Screen; Business Hypothesis Testing and a Simple Python Debug; Project Discussion and Open-Ended Case Study.
008: Data Scientist PhD Intern — phone or technical screen
Report date: 2026-02-04 · Role: Data Scientist PhD Intern · Stage: Phone / technical screen
Microsoft PhD Data Scientist internship report with an OA and three back-to-back interviews.
- Show an Organization Chart One Layer at a Time. Model the organization chart as a tree and show each complete layer from top to bottom. Breadth-first search with a queue is the expected traversal shape. Practice this problem
- Design Counterfactual Evaluation When A/B Testing Is Unavailable. Explain how to evaluate a change when an A/B test cannot be run. Discuss counterfactual-estimation approaches.
- Diagnose a Sudden Product-Metric Drop. Choose metrics for evaluating whether a product is good. If a core metric suddenly drops on one day, determine possible causes and how to test them.
Other reported topics or limits: Two Unnamed One-Hour OA Questions; Unnamed SQL Question; ML Experimentation and Evaluation Topics.
009: Data Scientist PhD Intern — phone or technical screen
Report date: 2026-01-05 · Role: Data Scientist PhD Intern · Stage: Phone / technical screen · Interview date: 2025-11-03
Microsoft Data Scientist PhD internship report with a HackerRank OA, three back-to-back interviews, LLM-heavy research context, and an offer.
- Implement a Precision-Recall Curve. Code or draw a precision-recall curve. The interviewer gave substantial hints after the candidate checked the definition of recall. Practice this problem
Other reported topics or limits: One Unnamed LeetCode and One Unnamed Data-Processing OA Task; KNN, PCA, and LLM Fundamentals.
010: Applied Scientist PhD Intern — phone or technical screen
Report date: 2025-12-20 · Role: Applied Scientist PhD Intern · Stage: Phone / technical screen · Interview date: 2025-11-20
Microsoft Seattle Applied Scientist Intern 2026 Summer report covering the application timeline, ML and LLM fundamentals, an unspecified easy Python string task, a rejection, and an author clarification that the candidate was a PhD student.
Other reported topics or limits: ML and LLM Foundations for Real-World Deployment; Unspecified Easy Python String Task.
011: Applied Scientist PhD Intern — phone or technical screen
Report date: 2025-11-10 · Role: Applied Scientist PhD Intern · Stage: Phone / technical screen
Microsoft Seattle Applied Scientist Intern 2026 Summer report covering an unusually direct technical first round, a deep research-project and ML evaluation discussion, two behavioral questions, advancement to a final round, and eventual rejection.
Other reported topics or limits: Deep Research-Project and ML Evaluation Discussion.
012: Applied Scientist PhD Intern — phone or technical screen
Report date: 2025-10-28 · Role: Applied Scientist PhD Intern · Stage: Phone / technical screen · Interview date: 2025-10-17
Microsoft Applied Scientist PhD internship report covering a 30-minute initial screen, three 45-minute final interviews, DS/LLM preparation mismatch, and an offer.
Other reported topics or limits: Explain How ChatGPT Is Trained; Three DS-Focused Problem-Solving and Coding Interviews.
014: Applied Scientist Intern — phone or technical screen
Report date: 2025-05-07 · Role: Applied Scientist Intern · Stage: Phone / technical screen · Interview date: 2024-11-01
Microsoft Applied Scientist internship report for the 2025 recruiting cycle, covering three coding interviews, resume questions, and behavioral discussion.
Other reported topics or limits: Two Unnamed Array Problems Using Two Pointers and a Sliding Window; Write an Unspecified String-Processing Function.
015: Applied Science Intern (Health & Life Sciences) — final interview loop
Report date: 2025-02-28 · Role: Applied Science Intern (Health & Life Sciences) · Stage: Final loop / onsite
Microsoft Health & Life Sciences Applied Science PhD internship final-round report from Montreal. Three research-talk interviews focused on the candidate's PhD work, several LLM/ML fundamentals, deep project challenges, and problem-solving experience.
Other reported topics or limits: Research Talk, LLM Fundamentals, and Project-Depth Discussion.
016: Student Researcher, LLM for Code — phone or technical screen
Report date: 2025-02-10 · Role: Student Researcher, LLM for Code · Stage: Phone / technical screen · Interview date: 2024-12-01
Microsoft LLM-for-code Student Researcher report describing a referral, two interviewers, resume and research discussion, one unnamed easy LeetCode task, a pass/offer path, and the author's clarification about part-time hourly pay.
Other reported topics or limits: Unspecified Easy LeetCode Task in a Research-Heavy Screen.
017: Data Scientist Intern — phone or technical screen
Report date: 2024-11-15 · Role: Data Scientist Intern · Stage: Phone / technical screen
Microsoft Data Scientist internship report from a CS PhD candidate, covering three back-to-back rounds, behavioral questions, one detailed median question with an optimization follow-up, and one Medium LeetCode task per round.
- Find a Median with a Memory-Constrained Follow-Up. Find a median using a min-heap and max-heap. When memory is constrained, use Quickselect as the follow-up approach.
Other reported topics or limits: Two Other Unnamed Medium LeetCode Questions.
018: Data Science PhD Intern — phone or technical screen
Report date: 2024-10-20 · Role: Data Science PhD Intern · Stage: Phone / technical screen · Interview date: 2024-10-01
Microsoft Data Science PhD internship report for the 2025 recruiting cycle, with three consecutive interviews covering coding, ML/probability, research, and LLM design, followed by an offer.
- Find the Median from Two Arrays. Given two arrays, find their median. The interviewer repeatedly asked for optimizations. Practice this problem
Other reported topics or limits: Expected Exit Time with Returning Paths; Bias–Variance, Linear Regression, Overfitting, and LLM Design.
019: AI Research Intern (non-MSR) — phone or technical screen
Report date: 2024-05-23 · Role: AI Research Intern (non-MSR) · Stage: Phone / technical screen · Interview date: 2024-05-17
Microsoft AI Research Intern pass report from a math PhD who moved into LLM reasoning and theorem proving.
- Unnamed LeetCode Easy Live-Coding Task. The second-round interview included one LeetCode Easy problem written and reviewed live. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Design How a Company Should Use AI; LLM Breadth and Research Discussion Set.
020: Research Intern — final interview loop
Report date: 2024-03-06 · Role: Research Intern · Stage: Final loop / onsite · Interview date: 2024-03-06
Detailed Microsoft product-oriented research-team summer-intern offer report. The first round aligned research interests and projects; the second round with an applied scientist included an ML-design question for the summer project; the third covered research experience and behavioral questions, with no coding test.
Other reported topics or limits: Design the ML Approach for an Unnamed Summer Project; Research Alignment, Technology Familiarity, and Behavioral Questions.
021: Mixed Reality Research Scientist Intern — phone or technical screen
Report date: 2024-01-16 · Role: Mixed Reality Research Scientist Intern · Stage: Phone / technical screen
Microsoft Mixed Reality Research Scientist Intern coding report with one concrete in-place word-order reversal task.
- Reverse Word Order In Place. Reverse the order of words in a string. The characters within each word remain unchanged. The interviewer required an in-place solution. Practice this problem
History (2020–Oct 2023)
These 98 reports are more than three years old. Most are Data Scientist and Data & Applied Scientist internships from 2020–2022, when final loops leaned on product and data-science cases. Their named questions:
Algorithms / Data Structures (1 of 3)
| Question | Reports | Last seen |
|---|---|---|
| Retain the Top 1,000 Values from a Large Integer Array | 2: 013, 112 | Oct 2022 |
| Is Graph Bipartite? | 2: 048, 066 | Feb 2022 |
| Search in Rotated Sorted Array | 1: 023 | Oct 2023 |
| Binary Tree Right Side View | 1: 023 | Oct 2023 |
| Compute a Median and Maintain It as Numbers Arrive | 1: 033 | Oct 2022 |
| Implement an LRU Cache | 1: 033 | Oct 2022 |
| Count Set Bits with a Precomputed Byte Lookup Table | 1: 013 | Oct 2022 |
| Wrap Text to 80 Characters Without Splitting Words | 1: 034 | Oct 2022 |
| Compute the Sum of Prime Divisors | 1: 042 | Mar 2022 (report date) |
| LeetCode 70, Climbing Stairs | 1: 042 | Mar 2022 (report date) |
| LeetCode 189, Rotate Array | 1: 044 | Feb 2022 |
| Reverse Words in a Sentence | 1: 048 | Feb 2022 |
| Convert a String to an Integer with Overflow Detection | 1: 048 | Feb 2022 |
Algorithms / Data Structures (2 of 3)
| Question | Reports | Last seen |
|---|---|---|
| Refactor a Regex-Based Noun Pluralization Function | 1: 051 | Dec 2021 |
| List All Files in a Folder and Find Duplicate Files | 1: 060 | Nov 2021 |
| Longest Increasing Subsequence | 1: 056 | Oct 2021 |
| Count Connected Components in a Grid | 1: 065 | Oct 2021 (report date) |
| Count Isolated Islands | 1: 066 | Oct 2021 |
| 3Sum | 1: 066 | Oct 2021 |
| LeetCode 1662, Check If Two String Arrays Are Equivalent | 1: 070 | Sep 2021 |
| Maximum Sum Increasing Subsequence | 1: 076 | Sep 2021 (report date) |
| Maximum Subarray Sum | 1: 077 | Mar 2021 (report date) |
| Find All Unique Two-Sum Pairs | 1: 077 | Mar 2021 (report date) |
| Find All Primes Less Than N | 1: 077 | Mar 2021 (report date) |
| Explain P-Values and Confidence Intervals | 1: 078 | Mar 2021 (report date) |
| Explain P-Values | 1: 082 | Feb 2021 |
Algorithms / Data Structures (3 of 3)
| Question | Reports | Last seen |
|---|---|---|
| Convert a Binary Search Tree to a Doubly Linked List | 1: 086 | Dec 2020 |
| Count Trailing Zeros in N Factorial | 1: 086 | Dec 2020 |
| Bold a Target Substring in an HTML Article | 1: 086 | Dec 2020 |
| Output Consecutive Numbers in an Array | 1: 089 | Dec 2020 (report date) |
| First Missing Positive | 1: 090 | Nov 2020 (report date) |
| Duplicate Zeros | 1: 095 | Nov 2020 |
| Generate a Fibonacci-Style Sequence with Constant Space and a Mathematical Solution | 1: 104 | Apr 2020 (report date) |
| Number of Islands | 1: 109 | Feb 2020 |
| Return the Top K Search-Query Suggestions for a Prefix | 1: 110 | Feb 2020 (report date) |
| Print Vertical Asterisk Bars | 1: 113 | Jan 2020 (report date) |
| Word Search with Eight-Direction Adjacency | 1: 113 | Jan 2020 (report date) |
| Knapsack Problem Variants | 1: 112 | Jan 2020 |
ML and statistics coding
| Question | Reports | Last seen |
|---|---|---|
| Implement a Precision-Recall Curve | 2: 056, 064 | Oct 2021 |
| Compute Variance in O(n) Time | 2: 078, 082 | Mar 2021 (report date) |
| Find the Nth Percentile, Including a Large Unsortable Input | 1: 033 | Oct 2022 |
| Estimate Pi Using Uniform[0,1] Samples | 1: 033 | Oct 2022 |
| Generate a Uniform Bounded Integer from Bernoulli(p) | 1: 033 | Oct 2022 |
| Generate N-Grams from a URL | 1: 046 | Feb 2022 |
| Two-Player Gambler's Ruin Probability and Simulation | 1: 060 | Nov 2021 |
| Enumerate Unique Unordered Dice Outcomes | 1: 056 | Oct 2021 |
| Count New-User Devices That Opened an App | 1: 067 | Oct 2021 |
| Score How Well a List Is Shuffled | 1: 066 | Oct 2021 |
| Percentage of Observed Users Among Installed Users by App | 1: 057 | Sep 2021 |
| Even-Valued Elements at Even Indices | 1: 079 | Feb 2021 |
| Implement Softmax Cross-Entropy | 1: 107 | Feb 2020 (report date) |
ML and system design (1 of 2)
| Question | Reports | Last seen |
|---|---|---|
| Design a Text Advertising Recommendation System | 2: 065, 109 | Oct 2021 (report date) |
| Design a System to Detect and Correct User String Typos | 1: 028 | Dec 2022 (report date) |
| Design Website Classification with Crowdsourced Label Quality Controls | 1: 033 | Oct 2022 |
| Design CTR Prediction from Query, Ad, and Click History | 1: 013 | Oct 2022 |
| Design User Targeting for Account, Subscription, and Data-Sharing Conversion | 1: 034 | Oct 2022 |
| Design a Query-to-Document Information Retrieval System with Click Feedback | 1: 041 | Apr 2022 (report date) |
| Design a Question-Answering Search Engine | 1: 044 | Feb 2022 |
| Design Related Query Suggestions for a Search Engine | 1: 059 | Nov 2021 |
| Design Reinforcement Learning for Topic Extraction | 1: 059 | Nov 2021 |
ML and system design (2 of 2)
| Question | Reports | Last seen |
|---|---|---|
| Design a Video Recommendation System from User, Video, and Action Tables | 1: 065 | Oct 2021 (report date) |
| Design Phone-Call Quality Classification from Limited User Feedback | 1: 063 | Oct 2021 |
| Design Webpage Language Detection and Language-Aware Search Ranking | 1: 066 | Oct 2021 |
| Design a Highway Toll Booth System | 1: 086 | Dec 2020 |
| Design Storage for Big Data | 1: 098 | Oct 2020 |
| Design a Ranking System | 1: 098 | Oct 2020 |
| Design a Self-Driving Car Perception System | 1: 103 | Sep 2020 (report date) |
| Design a Face-Recognition Neural Network Under Evolving Requirements | 1: 104 | Apr 2020 (report date) |
| Design an Excel Next-Command Recommender | 1: 109 | Feb 2020 |
ML, data science and product cases (1 of 2)
| Question | Reports | Last seen |
|---|---|---|
| Improve a Black-Box Predictive Model | 1: 013 | Oct 2022 |
| Represent a Directed Entity Network and Predict New Links | 1: 034 | Oct 2022 |
| Implement a Model Fine-Tuning Workflow | 1: 042 | Mar 2022 (report date) |
| Detect a Cycle in a Linked List | 1: 048 | Feb 2022 |
| Predict a Car's Price from Similar Cars | 1: 060 | Nov 2021 |
| Predict Product Ratings from User Reviews | 1: 065 | Oct 2021 (report date) |
| Clean and Tidy Patient Vitals Data | 1: 067 | Oct 2021 |
| Forecast Car-Model Sales and Evaluate a 2% Price Increase | 1: 067 | Oct 2021 |
| Evaluate Monthly Active Device Estimates Without Ground Truth | 1: 071 | Sep 2021 |
| Diagnose Bias in Product Feedback Data | 1: 078 | Mar 2021 (report date) |
| Formulate and Solve Cloud Resource Allocation | 1: 078 | Mar 2021 (report date) |
| Label a Million-Row Dataset from a Smaller Labeled Set | 1: 079 | Feb 2021 |
| Diagnose Bias in a Dataset | 1: 082 | Feb 2021 |
ML, data science and product cases (2 of 2)
| Question | Reports | Last seen |
|---|---|---|
| Detect Fake News from One Day of Tweets | 1: 089 | Dec 2020 (report date) |
| Use a Dataset to Help a Sales Department | 1: 093 | Nov 2020 (report date) |
| Quantify Pure Causal Effects for a Mobile Product | 1: 094 | Nov 2020 (report date) |
| Model What Drives Free-Trial Users to Purchase | 1: 094 | Nov 2020 (report date) |
| Diagnose and Model User Churn | 1: 095 | Nov 2020 |
| Match Anonymous Records from the Same Person | 1: 095 | Nov 2020 |
| Allocate Resources by Demand Density | 1: 095 | Nov 2020 |
| Evaluate a Windows Notification System | 1: 092 | Oct 2020 |
| Model User Preferences and Segment Customers for a Photo Product | 1: 092 | Oct 2020 |
| Classify Strings as Valid or Garbage | 1: 092 | Oct 2020 |
| Measure the Incremental Value of Desktop and Mobile Clients | 1: 096 | Oct 2020 (report date) |
| Implement a Direction-Specific Research Model | 1: 110 | Feb 2020 (report date) |
| Implement a Sequence-to-Sequence Model | 1: 110 | Feb 2020 (report date) |
013: Data Scientist / Research PhD Intern — phone or technical screen
Report date: 2025-06-18 · Role: Data Scientist / Research PhD Intern · Stage: Phone / technical screen · Interview date: 2022-10-10
The 2025 post is a retrospective of a September–October 2022 Microsoft PhD Data Scientist/Research internship loop.
- Retain the Top 1,000 Values from a Large Integer Array. Given a large integer array, find the top 1,000 values. Practice this problem
- Count Set Bits with a Precomputed Byte Lookup Table. Count the number of one bits in a byte from 0 through 255. Extend the approach to an array of one million numbers by precomputing results for all byte values. Practice this problem
- Design CTR Prediction from Query, Ad, and Click History. Given many historical queries, many ads, and click observations for each ad, frame an ML model that predicts click-through rate.
- Improve a Black-Box Predictive Model. Given training data and a black-box predictor f(X)=y, explain how to improve the model. Consider dimension reduction or feature selection, kernel methods or interactions, ensembles, personalized modeling, and data augmentation. Discuss which approaches can overfit.
Other reported topics or limits: Regression, CNN Parameters, and Attention; A/B-Test Sample Size and Exploration–Exploitation.
022: Data Scientist Intern — final interview loop
Report date: 2023-10-21 · Role: Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2023-10-01
Detailed Microsoft Data Scientist internship offer report.
- Unnamed SQL Medium Whiteboard Task. The first virtual-onsite round included one SQL Medium problem written on a whiteboard. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Two Contextual NLP/ML Cases; Recommendation Data, Metrics, and Customer-Segmentation Case; Experimentation, NLP Concepts, and Behavioral Questions.
023: Data Scientist Intern — final interview loop
Report date: 2023-10-14 · Role: Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2023-10-04
Microsoft Data Scientist Intern final-round report with three rounds.
- Search in Rotated Sorted Array. The original author identifies the first-round coding task as LeetCode 33. The author completed it quickly with one less-than-or-equal versus less-than bug. Practice this problem
- Binary Tree Right Side View. The original author identifies the third-round coding task as LeetCode 199. The first solution used DFS and the interviewer requested a BFS rewrite. Practice this problem
Other reported topics or limits: Time-Series Anomaly Detection Case.
024: Applied Scientist Intern — final interview loop
Report date: 2023-10-13 · Role: Applied Scientist Intern · Stage: Final loop / onsite · Interview date: 2023-10-12
Microsoft Applied Scientist Intern final-round report.
- Unspecified Applied Scientist Coding Question. The second final-round interview was a coding round. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Unspecified ML Design Round; Unspecified ML Theory Round.
025: Microsoft Research Intern — final interview loop
Report date: 2023-03-09 · Role: Microsoft Research Intern · Stage: Final loop / onsite · Interview date: 2023-02-08
Microsoft Research internship pass report.
- Unnamed Simple Coding Question. A senior interviewer asked one simple coding question during the follow-up round. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Candidate Research Talk and Proposal Discussion.
026: Office of Applied Research Intern — phone or technical screen
Report date: 2023-02-14 · Role: Office of Applied Research Intern · Stage: Phone / technical screen
Microsoft Office of Applied Research internship pass report. Three research scientists conducted two rounds about the candidate's research and project ideas; the post gives no coding or reusable design prompt.
Other reported topics or limits: Discuss Research and Project Ideas.
027: Microsoft Research Computational Biology Intern — final interview loop
Report date: 2023-02-10 · Role: Microsoft Research Computational Biology Intern · Stage: Final loop / onsite
Microsoft Research Health Futures computational-biology summer-intern report. The process moved from a principal-researcher screen to two senior-researcher interviews and a director conversation, centered on research projects, big-picture insight, ML, and generative-model knowledge.
Other reported topics or limits: Computational-Biology, ML, and Generative-Model Research Discussion.
028: Data Scientist / Applied Scientist PhD Intern — final interview loop
Report date: 2022-12-24 · Role: Data Scientist / Applied Scientist PhD Intern · Stage: Final loop / onsite
Detailed Microsoft Device-team Data Scientist/Applied Scientist PhD internship loop.
- Design a System to Detect and Correct User String Typos. Given a user-entered string, determine whether the user typed it incorrectly. If the string is wrong, produce or select the corrected string.
Other reported topics or limits: Unspecified Third-Round Case Study; Semi-Supervised Learning Case Study; Behavioral, Model-Detail, and DP-to-a-Child Discussion.
029: Research Intern — final interview loop
Report date: 2022-12-19 · Role: Research Intern · Stage: Final loop / onsite · Interview date: 2022-02-18
Concise Microsoft Research internship offer timeline. Two researcher conversations asked about ownership of prior implementation and the candidate's ideas for potential methods in an unnamed research area.
Other reported topics or limits: Propose a Potential Method for an Unnamed Research Problem; Implementation Ownership Discussion.
030: Applied Sciences Group Research Intern — final interview loop
Report date: 2022-12-01 · Role: Applied Sciences Group Research Intern · Stage: Final loop / onsite · Interview date: 2022-12-01
Microsoft Applied Sciences Group research-intern report from a third-year PhD in an AI-adjacent field.
- Unspecified Recursive Backtracking Combination Task. The coding task required recursive backtracking. The original author recalls that it was roughly about finding all combinations. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Compute a Bayesian Posterior from Error Rates; Research Talk and ML Breadth/Depth Discussion.
031: Data Scientist / Data & Applied Scientist Intern — final interview loop
Report date: 2022-11-29 · Role: Data Scientist / Data & Applied Scientist Intern · Stage: Final loop / onsite
Microsoft DSA internship final-round report.
- Unnamed LeetCode Easy Codility Task. The first round used Codility for one LeetCode Easy-level algorithm problem. The report did not preserve a reconstructable task contract.
- Unnamed SQL Task. The second final-round interviewer asked the candidate to write one SQL solution. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Unspecified Data-Science Case; Resume ML, High-Level DS, and Preference Discussion.
032: Data & Applied Sciences Intern — phone or technical screen
Report date: 2022-11-04 · Role: Data & Applied Sciences Intern · Stage: Phone / technical screen
Microsoft Data & Applied Sciences internship first-round phone screen. The conversation covered proudest project, teamwork difficulty, explaining ML, motivation for Microsoft and data science, coding skills, qualification, and ML algorithms used in prior projects.
Other reported topics or limits: Behavioral, ML Explanation, and Project-Algorithm Screen.
033: Data Scientist Intern — final interview loop
Report date: 2022-10-31 · Role: Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2022-10-31
Detailed Microsoft Bing Search Data Scientist internship offer loop.
- Find the Nth Percentile, Including a Large Unsortable Input. Given a list of integers, find the Nth percentile. Follow-up: the list is very large and cannot be sorted in full. Practice this problem
- Estimate Pi Using Uniform[0,1] Samples. Use samples from a Uniform[0,1] source to estimate the value of pi. Practice this problem
- Compute a Median and Maintain It as Numbers Arrive. Given a number list, compute its median. Follow-up: maintain the median while numbers are added one at a time. Practice this problem
- Generate a Uniform Bounded Integer from Bernoulli(p). Given access to Bernoulli(p), generate randint(a,b). The source gives randint(1,10) as an example. Practice this problem
- Implement an LRU Cache. Implement an LRU Cache. Practice this problem
- Design Website Classification with Crowdsourced Label Quality Controls. Design a classifier for websites. Discuss feature choices such as domain and page content. Labels come from crowdsourcing; address label quality and class imbalance. Estimate crowd quality by injecting examples with known labels and measuring correctness.
Other reported topics or limits: ML Fundamentals and Behavioral Questions.
034: Data Scientist and Applied Scientist Intern — final interview loop
Report date: 2022-10-11 · Role: Data Scientist and Applied Scientist Intern · Stage: Final loop / onsite · Interview date: 2022-10-10
Detailed Microsoft Data Scientist and Applied Scientist internship offer report.
- Wrap Text to 80 Characters Without Splitting Words. Given a very long text string, format it for printing with at most 80 characters per line. A line ending must not split a word. No output line may begin with a space. The interviewer asked for time and space complexity. Practice this problem
- Design User Targeting for Account, Subscription, and Data-Sharing Conversion. Design how to identify users likely to create a Microsoft account. Design how to identify existing account holders likely to purchase a subscription. Design how to persuade or identify users likely to consent to sharing data. The discussion considered device count, OneDrive capacity usage, and engagement frequency as candidate features. A follow-up asked how to use data to convince a product manager that the selected users are the right target.
- Represent a Directed Entity Network and Predict New Links. Entities form directed links such as A to B, B to C, C to D and A, and D to A and C. Choose a data structure to represent the network. Extend the representation to weighted edges and explain how to assign weights. When a new node arrives, predict which existing nodes it will connect to. Discuss whether outliers should be removed. The candidate proposed node embeddings and similarity for link prediction.
Other reported topics or limits: Python Scope, Deep-Learning Explanation, and Behavioral Questions.
035: Data & Applied Sciences PhD Intern — phone or technical screen
Report date: 2022-10-11 · Role: Data & Applied Sciences PhD Intern · Stage: Phone / technical screen · Interview date: 2022-10-11
Microsoft Data & Applied Sciences PhD internship phone-screen pass report. The 30-minute call covered project blockers, difficult collaboration, motivation, graduation timing and languages, explaining the candidate's research without technical terms, and explaining deep learning to a child. The author later reported an October 31 final round and November 8 offer but did not publish the final-round questions.
Other reported topics or limits: Behavioral, Research, and Deep-Learning Explanation Screen.
036: Data and Applied Scientist Intern — phone or technical screen
Report date: 2022-10-10 · Role: Data and Applied Scientist Intern · Stage: Phone / technical screen
Microsoft Data and Applied Scientist Intern screen report. The 40-minute conversation was almost entirely behavioral and applied-ML/process discussion, with no coding or reusable design deliverable; all questions remain valuable Experience-article evidence.
Other reported topics or limits: DAS Behavioral and Applied-ML Screen Question Set.
037: Data & Applied Sciences Intern — phone or technical screen
Report date: 2022-10-10 · Role: Data & Applied Sciences Intern · Stage: Phone / technical screen
Microsoft Data & Applied Sciences internship phone-screen report. The screen covered motivation for technology, favorite ML area, strongest programming language, explaining deep learning to a five-year-old, difficult teamwork, and overcoming a challenge. The author later reported no response and no later interview content.
Other reported topics or limits: Motivation, ML Preference, and Behavioral Screen.
038: Data & Applied Scientist Intern — phone or technical screen
Report date: 2022-10-07 · Role: Data & Applied Scientist Intern · Stage: Phone / technical screen · Interview date: 2022-09-26
Detailed Microsoft Data & Applied Scientist internship phone screen. Six reported questions covered collaboration difficulty, a challenging project and model choice, proudest work/impact, goals and obstacles, the generic data-science project process, and preferred DS area/team match.
Other reported topics or limits: Behavioral, Model-Choice, and DS-Process Screen.
039: Data & Applied Science Intern — phone or technical screen
Report date: 2022-09-22 · Role: Data & Applied Science Intern · Stage: Phone / technical screen · Interview date: 2022-09-20
Microsoft Data & Applied Science summer-intern phone-screen pass report. The half-hour call covered preferred DS areas, strongest programming language, a challenging project, explaining deep learning to a five-year-old, difficult teamwork, achieving a professional goal, favorite Microsoft product, and motivation for Microsoft. The candidate later received a final-round invitation but did not share those questions.
Other reported topics or limits: DS Preference, Project, and Behavioral Screen.
040: Microsoft Research HCI / Human+AI Intern — phone or technical screen
Report date: 2022-05-05 · Role: Microsoft Research HCI / Human+AI Intern · Stage: Phone / technical screen
MSR Montreal research-intern rejection report from a third-year HCI+AI PhD. Two principal researchers conducted three research-only rounds for HCI, computational social science, and human+AI groups. The candidate said there was no coding.
Other reported topics or limits: HCI and Human+AI Research Discussion.
041: Data & Applied Sciences Intern — phone or technical screen
Report date: 2022-04-27 · Role: Data & Applied Sciences Intern · Stage: Phone / technical screen
Detailed Microsoft Europe Data & Applied Sciences internship phone-screen report from an NLP master's candidate.
- Design a Query-to-Document Information Retrieval System with Click Feedback. Design an information-retrieval system in which a user submits a query and the system returns documents. Begin with the simplest viable idea and evolve it into a more mature design. Explain how observed user clicks can be used to improve the retrieval system.
Other reported topics or limits: Resume Classification-Model Evaluation Discussion.
042: Data Scientist Intern — final interview loop
Report date: 2022-03-17 · Role: Data Scientist Intern · Stage: Final loop / onsite
Microsoft Data Scientist internship virtual-onsite report.
- Compute the Sum of Prime Divisors. The coding task is described as computing the sum of prime divisors. Practice this problem
- Implement a Model Fine-Tuning Workflow. Choose a model and write a fine-tuning process. Documentation and external libraries were allowed. The report did not preserve a reconstructable task contract.
- LeetCode 70: Climbing Stairs. The third onsite coding task is explicitly identified as LeetCode 70. Practice this problem
Other reported topics or limits: Unnamed Prime, SQL, Probability, Statistics, CV, NLP, and ML Questions.
043: Data and Applied Scientist Summer Intern — phone or technical screen
Report date: 2022-02-25 · Role: Data and Applied Scientist Summer Intern · Stage: Phone / technical screen
Microsoft Data and Applied Scientist summer-intern phone-screen report. The interview covered internship history, a challenging project, private/public databases used, preferred ML/DL area, neural-network types, difficult teamwork, motivation, favorite Microsoft product, and explaining deep learning to a child. The author accepted another offer before a possible second round.
Other reported topics or limits: Database, ML Interest, Neural-Network, and Behavioral Screen.
044: Data & Applied Science Intern (Master's) — final interview loop
Report date: 2022-02-16 · Role: Data & Applied Science Intern (Master's) · Stage: Final loop / onsite · Interview date: 2022-02-03
Detailed Microsoft Data & Applied Science master's internship offer report for WebXT.
- Design a Question-Answering Search Engine. Design a Question Answering system similar to a Google search engine. Explain where the data comes from. Explain how to process the data. Design ranking. Define evaluation. Discuss additional improvements.
- LeetCode 189: Rotate Array. The fourth onsite algorithm is explicitly LeetCode 189 Rotate Array. The solution must be in place. The candidate used the double-reversal technique and had to explain why it works. Practice this problem
- Unnamed O(n)-Space to O(1)-Space, One-Pass Optimization Task. The initial solution used O(n) time and O(n) space. The interviewer guided an optimization to O(1) space and then one pass. The report did not preserve a reconstructable task contract.
- Unnamed Prefix-Sum Coding Task. The second onsite included an easy prefix-sum coding problem. The report did not preserve a reconstructable task contract.
- Unnamed Many-Queries Trie Task. The problem involved many queries and was solved with a trie. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Project, ML, Transformer, and Behavioral Questions.
045: Data & Applied Scientist Intern — final interview loop
Report date: 2022-02-15 · Role: Data & Applied Scientist Intern · Stage: Final loop / onsite · Interview date: 2022-02-11
Microsoft Data & Applied Scientist internship report. A short phone screen covered resume and behavioral topics; three initial virtual-onsite rounds plus a later added round mixed behavioral questions with SQL, Python, statistics, and ML. The added round presented a current team project and asked the candidate to explain how they would design it, but the project scenario and requirements were not shared.
Other reported topics or limits: Design an Unspecified Team ML Project; Unnamed SQL, Python, Statistics, ML, and Behavioral Questions.
046: Microsoft Research Security and AI Intern — phone or technical screen
Report date: 2022-02-08 · Role: Microsoft Research Security and AI Intern · Stage: Phone / technical screen · Interview date: 2022-02-03
Detailed Microsoft Research Security and AI internship offer report.
- Generate N-Grams from a URL. Given a URL, generate its n-grams in Python. The code had to be executed immediately in the candidate's own IDE. Follow-ups asked how to encode the n-grams, avoid a sparse-matrix problem, and handle extremely frequent and extremely rare tokens. Practice this problem
Other reported topics or limits: Research, Embedding, and Code-Analysis Discussion.
047: Data & Applied Scientist Intern — phone or technical screen
Report date: 2022-02-03 · Role: Data & Applied Scientist Intern · Stage: Phone / technical screen
Microsoft Data & Applied Scientist internship first-round report. An HR phone screen covered teamwork, learning new technologies, one resume project, and a brief technical discussion of ways to find outliers in a dataset.
Other reported topics or limits: Discuss Ways to Identify Outliers in a Dataset; Behavioral and Resume-Project Screen.
048: Data Scientist Intern (Bing Dev) — final interview loop
Report date: 2022-02-03 · Role: Data Scientist Intern (Bing Dev) · Stage: Final loop / onsite · Interview date: 2022-02-03
Microsoft Data Scientist Intern pass report for a Bing development team.
- Reverse Words in a Sentence. Reverse the order of words in a sentence. For This is it., the reported output is it. Is This.
- Convert a String to an Integer with Overflow Detection. Convert a string to an integer. Detect overflow (an input outside the 32-bit integer range) and throw an exception. Practice this problem
- Is Graph Bipartite?. Determine whether a graph is bipartite by coloring neighboring nodes with different colors. The original author confirms this was LeetCode 785 exactly. Practice this problem
- Detect a Cycle in a Linked List. Detect whether a linked list contains a cycle. Practice this problem
Other reported topics or limits: Behavioral, Project, Statistics, and Research Discussion.
049: Data & Applied Scientist / Data Scientist Intern — final interview loop
Report date: 2022-01-27 · Role: Data & Applied Scientist / Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2022-01-20
Microsoft Data & Applied Scientist internship offer report for the E&D MLX group. Four onsite rounds mixed deep resume/project probing, DL/NLP and ML fundamentals, low-difficulty algorithmic coding, and a detailed ML case study. The author explicitly withholds every concrete question and case scenario.
Other reported topics or limits: Withheld Detailed ML Case Study; Withheld Low-Difficulty Coding and ML/DL/NLP Questions.
050: Data & Applied Science Intern — final interview loop
Report date: 2022-01-26 · Role: Data & Applied Science Intern · Stage: Final loop / onsite · Interview date: 2022-01-20
Microsoft Data & Applied Science internship Hiring Event report centered on delayed results. The author states that four rounds covered statistics, ML, behavioral questions, experiment design, and a final database/SQL exercise, but publishes none of the individual questions or schemas.
Other reported topics or limits: Unnamed Statistics, ML, Experiment-Design, and Database Questions.
051: Data Scientist Intern — phone or technical screen
Report date: 2021-12-15 · Role: Data Scientist Intern · Stage: Phone / technical screen · Interview date: 2021-12-06
Detailed Microsoft Data Scientist internship phone-screen pass report.
- Refactor a Regex-Based Noun Pluralization Function. Explain an existing Python function that uses regular expressions to convert nouns to plural forms. Refactor or structurally improve that function. The report did not preserve a reconstructable task contract.
Other reported topics or limits: LightGBM Tuning, Early Stopping, and Explanation Questions.
052: Data & Applied Scientist Intern — phone or technical screen
Report date: 2021-11-30 · Role: Data & Applied Scientist Intern · Stage: Phone / technical screen · Interview date: 2021-11-30
Microsoft Data & Applied Scientist internship recruiter-screen report. The twenty-minute call contained only behavioral, project, motivation, team-preference, and ML/DL-interest questions, with no technical exercise or later-round update.
Other reported topics or limits: Leadership, Project, Motivation, and ML-Interest Screen.
053: Data Science / Applied Science Intern — phone or technical screen
Report date: 2021-11-29 · Role: Data Science / Applied Science Intern · Stage: Phone / technical screen · Interview date: 2021-11-29
Microsoft Data Science / Applied Science internship recruiter-screen report. The 30-minute first round asked five behavioral and research-communication questions; the author later received a virtual-onsite invitation but declined it after accepting another offer.
Other reported topics or limits: Behavioral and Research Communication Screen.
054: Microsoft Research AI Intern — final interview loop
Report date: 2021-11-23 · Role: Microsoft Research AI Intern · Stage: Final loop / onsite · Interview date: 2021-11-22
Microsoft Research AI summer-intern report. A manager project-direction chat led to two 30-minute interviews about the candidate's prior papers, detailed follow-ups, internship expectations, and preferred research topics.
Other reported topics or limits: Paper Follow-Ups and Research Preferences.
055: Data & Applied Science Intern — phone or technical screen
Report date: 2021-11-22 · Role: Data & Applied Science Intern · Stage: Phone / technical screen
Microsoft Data & Applied Science internship recruiter-screen report. The shortened holiday-season call covered pending offers, motivation for technology, a challenging data-science project, dataset size, leadership versus team-member role, a prior internship, motivation for Microsoft, preferred Microsoft area and product, and familiarity with deep learning. The author later reported a final-round invitation but no final-round details.
Other reported topics or limits: Project, Dataset, Team-Role, and Motivation Screen.
056: Data & Applied Scientist Intern — final interview loop
Report date: 2021-11-18 · Role: Data & Applied Scientist Intern · Stage: Final loop / onsite · Interview date: 2021-10-28
Detailed Microsoft Data & Applied Scientist internship pass report.
- Implement a Precision-Recall Curve. The inputs are a prediction array and a label array. Compute the precision and recall arrays that form the precision-recall curve. Practice this problem
- Longest Increasing Subsequence. The interviewer used LeetCode 300 exactly. The candidate wrote a dynamic-programming solution; the interviewer did not require the O(n log n) follow-up. Practice this problem
- Enumerate Unique Unordered Dice Outcomes. A die has k faces and is rolled n times. Output every possible result without duplicates. Outcomes such as (1,2) and (2,1) are considered the same. DFS met the initial expectation; larger n and k caused stack overflow and prompted a dynamic-programming follow-up. Practice this problem
Other reported topics or limits: Predict Windows Upgrade Duration; Satellite Cloud Detection and ML Fundamentals.
057: Data & Applied Scientist PhD Intern — phone or technical screen
Report date: 2021-11-11 · Role: Data & Applied Scientist PhD Intern · Stage: Phone / technical screen · Interview date: 2021-09-15
Detailed Microsoft Data & Applied Scientist PhD internship offer report.
- Percentage of Observed Users Among Installed Users by App. Table 1 has columns app and installed_users. Table 2 has columns app, userId, and date. For each app, compute the percentage of users represented in Table 2 among all installed users for that app. The candidate wrote SQL in a shared Word window during the Teams call. Practice this problem
Other reported topics or limits: High-Dimensional ML, Model Comparison, and Product Experimentation Questions; Unnamed Onsite ML Case Studies and Coding Questions.
058: Data & Applied Science / Data Science Intern — final interview loop
Report date: 2021-11-09 · Role: Data & Applied Science / Data Science Intern · Stage: Final loop / onsite · Interview date: 2021-10-28
Microsoft Data & Applied Science internship offer timeline. The Azure virtual onsite had two coding-plus-behavioral rounds, a statistics and machine-learning fundamentals round, and a final behavioral-plus-simple-coding round. The author characterizes coding as ordinary algorithms/data structures around medium difficulty but discloses no individual task identity.
Other reported topics or limits: Unnamed Azure Algorithms, Data Structures, Statistics, and ML Questions.
059: Data Science PhD Intern — final interview loop
Report date: 2021-11-07 · Role: Data Science PhD Intern · Stage: Final loop / onsite · Interview date: 2021-11-02
Microsoft Data Science PhD internship final-round report.
- Design Related Query Suggestions for a Search Engine. A search engine should return relevant search suggestions for a user's query. Propose an algorithm and explain its details.
- Design Reinforcement Learning for Topic Extraction. Explain how reinforcement learning could be applied to topic extraction.
Other reported topics or limits: Four Unnamed Easy-to-Medium Onsite Coding Tasks.
060: Data Scientist Intern — final interview loop
Report date: 2021-11-04 · Role: Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2021-11-01
Detailed Microsoft Data Scientist internship onsite report.
- Predict a Car's Price from Similar Cars. Write code to predict car prices. The candidate used nearest neighbors and averaged their prices.
- List All Files in a Folder and Find Duplicate Files. Print every file in a folder. Follow-up: find files that are the same. Follow-up: optimize the duplicate-file search. Practice this problem
- Two-Player Gambler's Ruin Probability and Simulation. Two people gamble against a machine. A win changes wealth by plus one and a loss by minus one. Both people start with one unit of wealth. Compute the probability that both eventually lose all wealth. After deriving the probability, write and run a simulation. Practice this problem
Other reported topics or limits: Predict Unspecified User Demand; Unnamed General DS Knowledge and Table-Merge Questions.
061: Data & Applied Scientist PhD Intern — final interview loop
Report date: 2021-11-04 · Role: Data & Applied Scientist PhD Intern · Stage: Final loop / onsite · Interview date: 2021-10-28
Microsoft Data & Applied Scientist PhD internship offer report. This atypical four-round onsite consisted almost entirely of behavioral conversation and deep discussion of one candidate-specific project, including packages, storage, platform, and design rationale. The author confirms no coding round and no reusable standalone question.
Other reported topics or limits: Candidate-Specific End-to-End Project Deep Dive.
062: Data & Applied Science Intern / Data Scientist Intern — final interview loop
Report date: 2021-11-03 · Role: Data & Applied Science Intern / Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2021-09-30
Microsoft Data & Applied Science internship pass report that transferred to Data Scientist Intern. It preserves the complete recruiting timeline and states that the first round covered resume projects, basic ML knowledge, and behavioral questions, but intentionally withholds specific technical prompts.
Other reported topics or limits: Withheld ML, Resume-Project, and Behavioral Interview Content.
063: Data & Applied Science Summer Intern — final interview loop
Report date: 2021-10-27 · Role: Data & Applied Science Summer Intern · Stage: Final loop / onsite · Interview date: 2021-10-14
Detailed Microsoft Data & Applied Science summer-intern offer report.
- Design Phone-Call Quality Classification from Limited User Feedback. Determine whether a phone call's quality is good or poor. Only a small subset of calls has user-feedback labels. Describe the solution and write pseudocode covering train/test splitting and encoding.
- Unnamed Medium Sliding-Window Coding Task. The final onsite coding problem was an unfamiliar LeetCode Medium sliding-window problem. The report did not preserve a reconstructable task contract.
Other reported topics or limits: ML, Statistics, and Ensemble Fundamentals.
064: Data Scientist Intern — final interview loop
Report date: 2021-10-26 · Role: Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2021-10-21
Microsoft Data Scientist Intern final-round pass report for a search team.
- Implement a Precision-Recall Curve. The interviewer asked the candidate to code a precision/recall curve and discuss an optimal solution. The original author clarifies that the task required a simple manual implementation without library calls. The implementation did not need to render a graph; it only needed to demonstrate the approach and optimal algorithm. Practice this problem
Other reported topics or limits: Rank Passages and Choose a Bing Fact; Search and ML Concept Question Set.
065: Data & Applied Scientist Intern — final interview loop
Report date: 2021-10-23 · Role: Data & Applied Scientist Intern · Stage: Final loop / onsite
Detailed Microsoft Data & Applied Scientist master's internship onsite report for Bing Ads.
- Count Connected Components in a Grid. The task was described as similar to LeetCode 542. Return the number of connected components in a grid.
- Design a Text Advertising Recommendation System. Design a recommendation system for text advertisements. Cover feature construction, model choice, large-scale candidate retrieval, ranking, current ranking models, and ranking features.
- Design a Video Recommendation System from User, Video, and Action Tables. Design a video recommendation system from user, video, and action tables. Discuss one-hot and target encoding, evolving user interests, sliding-window features and window tuning, regression versus classification, and negative sampling.
- Predict Product Ratings from User Reviews. Given tables of user reviews, build a model to predict product ratings from one to five. Choose regression or classification; specify RNN inputs, outputs, and loss; compare RNN, GRU, and LSTM; and propose a non-deep-learning TF-IDF approach with high-dimensional sparse-feature regularization.
Other reported topics or limits: Tree, GBDT, Time-Series, Causal-Inference, and NLP Theory Follow-Ups; Two Dice Probability Decision Questions.
066: Data & Applied Scientist / Applied Scientist Intern — final interview loop
Report date: 2021-10-22 · Role: Data & Applied Scientist / Applied Scientist Intern · Stage: Final loop / onsite · Interview date: 2021-10-14
Detailed Microsoft Data & Applied Scientist internship loop for Bing Search, later transferred to Applied Scientist Intern.
- Count Isolated Islands. The phone screen asked a simple isolated-islands coding problem. Practice this problem
- Determine Whether an Adjacency-List Graph Is Bipartite. Given an adjacency list, determine whether the graph is bipartite. The expected approach used breadth-first search. Practice this problem
- Score How Well a List Is Shuffled. Given a list, produce a measure of how well it is shuffled. The candidate proposed the number of monotonic sequences divided by total length. A follow-up asked for a better method, and the candidate suggested considering distances between values in monotonic sequences. Practice this problem
- 3Sum. The final coding task was 3Sum. Practice this problem
- Design Webpage Language Detection and Language-Aware Search Ranking. Determine which language a webpage uses. Use the detected-language information when ranking search results.
Other reported topics or limits: ML/DL, Knowledge Graph, and Research Discussion.
067: Data & Applied Science PhD Intern — final interview loop
Report date: 2021-10-21 · Role: Data & Applied Science PhD Intern · Stage: Final loop / onsite · Interview date: 2021-10-21
Detailed Microsoft Data & Applied Science PhD internship offer report.
- Clean and Tidy Patient Vitals Data. A sample dataset contains patient ID, blood pressure, gender, heart rate, and similar fields. Explain and implement how to clean the data. Handle missing values. Split a blood-pressure value such as 120/80 into two variables. Normalize equivalent gender values such as F and Female. Discuss tidy-data structure and multiple missing-value strategies.
- Count New-User Devices That Opened an App. Table 1 contains device_id, opened_app, and open_count; device_id can repeat. Table 2 contains device_id, isNew, and additional columns. Find how many device IDs that opened an app belong to new users. Practice this problem
- Forecast Car-Model Sales and Evaluate a 2% Price Increase. A car dealer wants to predict which models will sell well and which will not. Discuss data processing, candidate models, and why to choose them. Follow-up: decide whether a two-percent price increase is viable and explain how to test or operate it.
Other reported topics or limits: Infer Windows Developer Feature Preferences Across Machine Types; Statistics, Regression, Ensemble, and Tree Fundamentals.
068: Data & Applied Science Intern — phone or technical screen
Report date: 2021-10-21 · Role: Data & Applied Science Intern · Stage: Phone / technical screen
Microsoft Data & Applied Science internship recruiter-screen report. The rapid call covered proudest and most challenging projects, motivation, preferred DS/ML area, data workflow, difficult teamwork, DS experience, NLP familiarity, programming language, a built deep-learning model, explaining deep learning to a child, and accommodating user needs. No later interview update was added.
Other reported topics or limits: Project, Data Workflow, NLP/DL, and Behavioral Screen.
069: Data/Applied Scientist Intern — phone or technical screen
Report date: 2021-10-13 · Role: Data/Applied Scientist Intern · Stage: Phone / technical screen · Interview date: 2021-10-13
Detailed Microsoft Data/Applied Scientist internship phone screen. The technical portion asked for verbal explanations of overfitting remedies, Ridge versus Lasso and variable selection, binary-classification metrics, and bagging versus boosting, followed by behavioral questions. No reusable implementation or design artifact was requested.
Other reported topics or limits: Overfitting, Regularization, Classification Metrics, and Ensemble Theory Screen.
070: Data & Applied Science PhD Intern — final interview loop
Report date: 2021-10-02 · Role: Data & Applied Science PhD Intern · Stage: Final loop / onsite · Interview date: 2021-09-23
Detailed Microsoft Data & Applied Science PhD internship offer report.
- LeetCode 1662: Check If Two String Arrays Are Equivalent. Determine whether two lists of string fragments represent the same word when concatenated. Example: ['Mic','ro','soft'] and ['M','ic','roso','ft'] both represent Microsoft. The candidate could use any language and run the code. Practice this problem
Other reported topics or limits: Diagnose a Metric Drop, Build the Dataset, and Choose a Classification Threshold; Unnamed Simple SQL and Statistics/ML Fundamentals.
071: Data & Applied Science Intern — final interview loop
Report date: 2021-09-30 · Role: Data & Applied Science Intern · Stage: Final loop / onsite · Interview date: 2021-09-23
Microsoft Data & Applied Science summer-intern offer report.
- Evaluate Monthly Active Device Estimates Without Ground Truth. A model estimates monthly active devices. There is no direct ground-truth label for the target metric. Design how to evaluate the model's predictions.
Other reported topics or limits: Identify Required Data and Write a Hypothetical SQL Query; Naive Bayes Spam Explanation and Behavioral Questions.
072: Data Scientist PhD Intern — phone or technical screen
Report date: 2021-09-30 · Role: Data Scientist PhD Intern · Stage: Phone / technical screen
Microsoft Data Scientist PhD internship first-round screen. The candidate received behavioral and background questions plus a child-friendly explanation of deep learning. There was no coding. The candidate later confirmed that the second round was the final round but did not share its questions.
Other reported topics or limits: Behavioral, Research, and Child-Friendly Deep Learning Screen.
073: Data & Applied Science PhD Intern — final interview loop
Report date: 2021-09-29 · Role: Data & Applied Science PhD Intern · Stage: Final loop / onsite · Interview date: 2021-09-23
Detailed Microsoft Data & Applied Science PhD internship offer report. The initial screen covered broad background and ML communication questions. Four onsite rounds were team-specific case studies requiring end-to-end analysis, data acquisition, modeling, and evaluation, plus one undisclosed SQL task. The candidate confirmed there was no coding.
Other reported topics or limits: Four Undisclosed Team-Project Case Studies; One Undisclosed SQL Task; Background, ML Communication, and Product Opinion Screen.
074: Data & Applied Scientist Intern — phone or technical screen
Report date: 2021-09-29 · Role: Data & Applied Scientist Intern · Stage: Phone / technical screen · Interview date: 2021-09-28
Microsoft Data & Applied Scientist Intern first-round report from a third-year PhD. The 30-minute Teams interview contained ten minutes of behavioral questions and a rapid ML-concept round with no coding.
Other reported topics or limits: Rapid ML Concepts and Behavioral Screen.
075: Data & Applied Scientist Intern — phone or technical screen
Report date: 2021-09-28 · Role: Data & Applied Scientist Intern · Stage: Phone / technical screen
Microsoft Data & Applied Scientist Intern first-round report. The 30-minute screen contained resume, collaboration, and impact questions followed by basic statistics and ML-concept discussion, with no coding exercise.
Other reported topics or limits: Behavioral and Basic ML Concept Screen.
076: Data/Applied Scientist Intern — final interview loop
Report date: 2021-09-15 · Role: Data/Applied Scientist Intern · Stage: Final loop / onsite
Detailed Microsoft Data/Applied Scientist internship report for an NLP group.
- Maximum Sum Increasing Subsequence. Given an array, consider all increasing subsequences. Return the maximum possible subsequence sum. Practice this problem
Other reported topics or limits: Array Median with Restricted Operations; Undisclosed NLP Case Study; Candidate-Specific Deep Learning Project Follow-Ups.
077: Data Scientist PhD Intern — final interview loop
Report date: 2021-03-20 · Role: Data Scientist PhD Intern · Stage: Final loop / onsite
Detailed Microsoft Data Scientist PhD internship offer report for a Bing Maps placement.
- Maximum Subarray Sum. Given an array, find the maximum sum of a contiguous subarray. Practice this problem
- Find All Unique Two-Sum Pairs. Solve a Two Sum variant. Return every distinct pair. Practice this problem
- Find All Primes Less Than N. Given n, return all prime numbers less than n. Practice this problem
Other reported topics or limits: Compare Two Companies' Similar Apps; One Forgotten Easy Coding Task; Behavioral, Project, ML, Data-Structure, and SQL Knowledge Discussion.
078: Applied/Data Scientist PhD Intern — final interview loop
Report date: 2021-03-01 · Role: Applied/Data Scientist PhD Intern · Stage: Final loop / onsite
Detailed Microsoft Applied/Data Scientist PhD internship onsite report.
- Compute Variance in O(n) Time. Given numerical observations, compute their variance. The required time complexity is O(n). Practice this problem
- Diagnose Bias in Product Feedback Data. Given a dataset, determine whether it is biased. For low feedback scores, distinguish product-caused dissatisfaction from other causes.
- Formulate and Solve Cloud Resource Allocation. Formulate a cloud-computing resource-allocation problem. Explain how to solve the resulting formulation.
Other reported topics or limits: Explain P-Values and Confidence Intervals.
079: Data Scientist Intern — phone or technical screen
Report date: 2021-02-25 · Role: Data Scientist Intern · Stage: Phone / technical screen · Interview date: 2021-02-25
Detailed Microsoft Data Scientist internship technical screen.
- Label a Million-Row Dataset from a Smaller Labeled Set. Dataset A contains approximately 100,000 labeled rows. Dataset B contains millions of rows and needs labels. Propose an effective labeling solution.
- Even-Valued Elements at Even Indices. Given a list, return every element whose index is even and whose value is also even. Practice this problem
Other reported topics or limits: Resume Project and Job-Description Match Discussion.
080: Microsoft Research NLP Intern — final interview loop
Report date: 2021-02-19 · Role: Microsoft Research NLP Intern · Stage: Final loop / onsite
Microsoft Research NLP internship video interview. Roughly twenty minutes focused on the candidate's research, followed by about ten minutes discussing an algorithm described only as similar to 3Sum and its complexity. The exact variant was not shared.
Other reported topics or limits: Unspecified 3Sum-Like Algorithm; Candidate-Specific Research Deep Dive.
081: Data Science PhD Intern — phone or technical screen
Report date: 2021-02-16 · Role: Data Science PhD Intern · Stage: Phone / technical screen · Interview date: 2021-02-16
Detailed Microsoft Data Science PhD internship background screen. The interviewer asked about ML interests, publications, research collaboration, proudest work, programming language, resume items, collaboration difficulty, explaining deep learning to a child, and a preferred Microsoft product. No reusable standalone artifact was requested.
Other reported topics or limits: ML Interests, Publications, Collaboration, and Product Screen.
082: Applied & Data Scientist PhD Intern — final interview loop
Report date: 2021-02-04 · Role: Applied & Data Scientist PhD Intern · Stage: Final loop / onsite · Interview date: 2021-02-04
Microsoft Applied & Data Scientist PhD internship onsite report.
- Compute Variance. Write code to compute variance. Practice this problem
- Diagnose Bias in a Dataset. Explain how to determine whether a dataset is biased.
Other reported topics or limits: Explain P-Values.
083: Microsoft Research Intern — phone or technical screen
Report date: 2020-12-18 · Role: Microsoft Research Intern · Stage: Phone / technical screen · Interview date: 2020-12-18
Microsoft Research internship report from a fourth-year PhD. Two one-hour interviews were entirely research-focused: the candidate presented recent work and answered detailed follow-ups about it. There was no coding or general technical-question round. The candidate later reported receiving an offer on December 29.
Other reported topics or limits: Present Recent Research and Answer Detail Questions.
084: Data & Applied Science PhD Intern — phone or technical screen
Report date: 2020-12-16 · Role: Data & Applied Science PhD Intern · Stage: Phone / technical screen · Interview date: 2020-12-16
Detailed Microsoft Data & Applied Science PhD internship first-round screen for the 2021 internship cycle. Questions covered ML interests, projects, publications, collaboration, programming languages, explaining deep learning to a child, Microsoft product improvement, motivation, and team preference. No reusable standalone artifact was requested.
Other reported topics or limits: ML Interests, Projects, Publications, Collaboration, and Product Screen.
085: Data Scientist Intern — phone or technical screen
Report date: 2020-12-14 · Role: Data Scientist Intern · Stage: Phone / technical screen · Interview date: 2020-12-08
Detailed Microsoft Data Scientist master's internship first round. The NLP-team interviewer spent about twenty-five minutes on one resume project, communication difficulty, model validation, and explaining p-values to a non-statistician. The author explicitly reports no coding and no case study.
Other reported topics or limits: Resume Project, Communication, Validation, and P-Value Screen.
086: Applied & Data Scientist Intern — final interview loop
Report date: 2020-12-09 · Role: Applied & Data Scientist Intern · Stage: Final loop / onsite · Interview date: 2020-12-09
Detailed Microsoft Applied & Data Scientist master's internship onsite.
- Convert a Binary Search Tree to a Doubly Linked List. Convert a binary search tree into a doubly linked list. Practice this problem
- Count Trailing Zeros in N Factorial. Given n, count the number of trailing zeros in n factorial. Practice this problem
- Bold a Target Substring in an HTML Article. The input is an HTML article and start and end indices for a target string. Modify the HTML so the target string appears bold. Practice this problem
- Design a Highway Toll Booth System. Design a highway toll booth on a whiteboard. Identify interacting components. Write high-level pseudocode for classes and functions.
Other reported topics or limits: XGBoost and Recommendation/NLP Improvement Discussions.
087: Data Scientist Intern — phone or technical screen
Report date: 2020-12-06 · Role: Data Scientist Intern · Stage: Phone / technical screen · Interview date: 2020-12-04
Microsoft Data Scientist internship first-round update. The author characterizes the round as behavioral questions plus simple algorithms, but only discloses motivation and candidate-specific project/challenge prompts; no algorithm identity or task details are given.
Other reported topics or limits: Motivation, Project Challenge, and Data-Science Problem Questions; Undisclosed Simple Algorithm Questions.
088: Data & Applied Science Intern — phone or technical screen
Report date: 2020-12-05 · Role: Data & Applied Science Intern · Stage: Phone / technical screen · Interview date: 2020-12-04
Detailed Microsoft Data & Applied Science master's internship first round. At least half of the interview covered motivation and candidate-specific project challenges, followed by one easy LeetCode-style Python task that the author solved using enumerate but never disclosed. A later update confirms a virtual-onsite invitation without adding questions.
Other reported topics or limits: Motivation, Project Challenge, and General Data-Science Problem Screen; Undisclosed Easy LeetCode-Style Python Task Using Enumerate.
089: Applied Data and Science PhD Intern — phone or technical screen
Report date: 2020-12-02 · Role: Applied Data and Science PhD Intern · Stage: Phone / technical screen
Detailed Microsoft Applied Data and Science PhD internship first-round interview for the 2021 cycle.
- Detect Fake News from One Day of Tweets. Given one day's tweets, design an approach to identify fake news. The interview discussion may consider signals such as positive versus negative comments.
- Output Consecutive Numbers in an Array. Given an array, output its consecutive numbers. Practice this problem
Other reported topics or limits: Text-Mining Project and Difficult Teammate Discussion.
090: Data Scientist PhD Intern — phone or technical screen
Report date: 2020-11-18 · Role: Data Scientist PhD Intern · Stage: Phone / technical screen
Detailed Microsoft Data Scientist PhD internship screen.
- First Missing Positive. Given an integer array, return the smallest missing positive integer. The task matches LeetCode 41. Practice this problem
Other reported topics or limits: Area Interest and Child-Friendly Internet Explanation.
091: Data & Applied Science PhD Intern — phone or technical screen
Report date: 2020-11-16 · Role: Data & Applied Science PhD Intern · Stage: Phone / technical screen
Detailed Microsoft Data & Applied Science PhD internship first-round screen. The interview covered self-introduction, NLP experience, a challenging project and model evaluation, explaining deep learning to a child, motivation, and Microsoft product improvement. Later posts only discuss delayed scheduling and add no interview question.
Other reported topics or limits: NLP, Project Evaluation, Motivation, and Product Screen.
092: Applied & Data Scientist PhD Intern — final interview loop
Report date: 2020-11-15 · Role: Applied & Data Scientist PhD Intern · Stage: Final loop / onsite · Interview date: 2020-10
Detailed Microsoft Applied & Data Scientist PhD internship offer report.
- Evaluate a Windows Notification System. Design an end-to-end evaluation for Windows system notifications. Use business judgment to choose features and metrics and propose a model. Expect follow-ups based on the proposed approach.
- Model User Preferences and Segment Customers for a Photo Product. For a photo-like product, explain how to understand user preferences. Design a customer-segmentation approach.
- Classify Strings as Valid or Garbage. Build a statistical model that classifies a string as valid or garbage. Cover the process from data collection through the statistical model. Discuss hypothesis testing and probability within the case.
Other reported topics or limits: Screen and Onsite Behavioral or Conceptual Questions.
093: Data Scientist PhD Intern — final interview loop
Report date: 2020-11-13 · Role: Data Scientist PhD Intern · Stage: Final loop / onsite
Concise Microsoft Data Scientist PhD internship virtual-onsite report.
- Use a Dataset to Help a Sales Department. Given a dataset, propose what analyses or models can be built. Explain how a data scientist could use it to help a sales department.
Other reported topics or limits: Undisclosed SQL Table Manipulation; Regularization, Missing Values, and Behavioral Discussion.
094: Data & Applied Science PhD Intern — final interview loop
Report date: 2020-11-09 · Role: Data & Applied Science PhD Intern · Stage: Final loop / onsite
Detailed Microsoft Data & Applied Science PhD internship offer report.
- Quantify Pure Causal Effects for a Mobile Product. Choose a frequently used mobile application and answer high-level product-analytics questions about it. Explain how to quantify pure causal effects, using ideas such as A/B testing or causal inference.
- Model What Drives Free-Trial Users to Purchase. Determine what drives users from free trials to purchase. Formulate the task as classification and consider logistic regression. Specify the data and answer model follow-ups such as regularization.
Other reported topics or limits: Screen Behavioral and ML Background Questions; Unspecified A/B Testing and Easy SQL Items.
095: Data & Applied Scientist PhD Intern — final interview loop
Report date: 2020-11-03 · Role: Data & Applied Scientist PhD Intern · Stage: Final loop / onsite · Interview date: 2020-11-02
Detailed Microsoft Data & Applied Scientist PhD internship virtual onsite.
- Duplicate Zeros. The coding task was LeetCode 1089, Duplicate Zeros. Practice this problem
- Diagnose and Model User Churn. Given a scenario in which users churn, explain why they are leaving. Develop the analysis strategy step by step and discuss the proposed model's pros and cons.
- Match Anonymous Records from the Same Person. Given anonymized data, determine which records originate from the same person.
- Allocate Resources by Demand Density. Allocate resources according to the density of demand. Define the target, how it will be measured, and how the allocation will be optimized.
Other reported topics or limits: Phone-Screen Probability, P-Value, and Forgotten Product Question; Undisclosed SQL Follow-Up After Duplicate Zeros.
096: Data & Applied Sciences PhD Intern — phone or technical screen
Report date: 2020-10-27 · Role: Data & Applied Sciences PhD Intern · Stage: Phone / technical screen
Detailed Microsoft Data & Applied Sciences PhD internship screen for the 2021 cycle.
- Measure the Incremental Value of Desktop and Mobile Clients. A Microsoft product is available on the web, a desktop client, and a mobile client. Determine whether the desktop and mobile clients create incremental product value. Specify the experimental procedure and explain statistical testing and p-values.
Other reported topics or limits: Resume and PhD Experience Discussion.
097: Data & Applied Science Intern — phone or technical screen
Report date: 2020-10-15 · Role: Data & Applied Science Intern · Stage: Phone / technical screen · Interview date: 2020-10-13
Microsoft Data & Applied Science internship phone screen. The six reported prompts covered prior ML projects, improving a favorite Microsoft product, team preference, explaining two familiar deep-learning models in depth, a difficult teamwork experience, and explaining AI to a five-year-old.
Other reported topics or limits: Behavioral, Product-Preference, and ML-Explanation Screen.
098: Data & Applied Scientist PhD Intern — phone or technical screen
Report date: 2020-10-06 · Role: Data & Applied Scientist PhD Intern · Stage: Phone / technical screen · Interview date: 2020-10-06
Detailed Microsoft Data & Applied Scientist PhD internship first-round video interview.
- Design Storage for Big Data. Explain how to handle storage for a large data workload.
- Design a Ranking System. Explain how to build a ranking system.
Other reported topics or limits: Foundational ML Concept Questions; Talk About Microsoft Products.
099: Data & Applied Sciences PhD Intern — phone or technical screen
Report date: 2020-09-29 · Role: Data & Applied Sciences PhD Intern · Stage: Phone / technical screen
Detailed Microsoft Data & Applied Sciences PhD internship recruiter screen. Thirteen reported prompts covered DS-versus-ML preference, project depth, coding and language background, data types, NLP/deep-learning model experience, teamwork, explaining deep learning to a child, publications, technology passion, product improvement, motivation, and team preference.
Other reported topics or limits: Recruiter Background, Behavioral, and Product-Preference Screen.
100: Data Science PhD Intern — phone or technical screen
Report date: 2020-09-28 · Role: Data Science PhD Intern · Stage: Phone / technical screen · Interview date: 2020-09-28
Detailed Microsoft Data Science PhD internship first-round screen for the 2021 cycle. The questions covered ML/NLP/DL projects, project difficulty and pride, teamwork, model evaluation, explaining deep learning to a child, interests, motivation, product improvement, and team preference. No reusable standalone artifact was requested.
Other reported topics or limits: ML, NLP, DL, Project, Teamwork, and Product Screen; Speculative Undisclosed Second-Round Topics.
101: Data Science PhD Intern — phone or technical screen
Report date: 2020-09-25 · Role: Data Science PhD Intern · Stage: Phone / technical screen · Interview date: 2020-09-25
Detailed Microsoft data-science PhD internship first-round phone screen. The questions covered ML background, research, behavioral examples, product preference, motivation, and team fit rather than concrete technical tasks.
Other reported topics or limits: General ML Background, Behavioral, and Product Screen.
102: Microsoft Research Redmond Machine Learning Intern — phone or technical screen
Report date: 2020-09-23 · Role: Microsoft Research Redmond Machine Learning Intern · Stage: Phone / technical screen
process summary from a returning Microsoft Research Redmond intern working on graph neural networks. Both interviews were research conversations about the candidate's papers and group projects, with no coding exercise.
Other reported topics or limits: Paper and Group-Project Research Discussion.
103: Data & Applied Sciences PhD Intern — phone or technical screen
Report date: 2020-09-16 · Role: Data & Applied Sciences PhD Intern · Stage: Phone / technical screen
Detailed Microsoft Data & Applied Sciences PhD internship technical phone screen.
- Design a Self-Driving Car Perception System. Design a perception system for a self-driving car. Formulate the task as a machine-learning problem. Specify the training data and evaluation metric. Explain how to handle imbalanced data. Discuss how the approach changes when detecting bicycles or trucks instead of cars.
Other reported topics or limits: Self-Introduction, Internship, and Research Background.
104: Machine Learning Intern — final interview loop
Report date: 2020-04-24 · Role: Machine Learning Intern · Stage: Final loop / onsite
Detailed Microsoft Machine Learning internship offer report.
- Generate a Fibonacci-Style Sequence with Constant Space and a Mathematical Solution. Generate a sequence in which each third/current value is the sum of the previous two values. The candidate gave a two-pointer constant-space solution and also a mathematical solution. Practice this problem
Other reported topics or limits: Design a Face-Recognition Neural Network Under Evolving Requirements; ML Theory, Research, Probability, and Behavioral Discussion.
105: Data Scientist Intern, Yammer team — phone or technical screen
Report date: 2020-04-19 · Role: Data Scientist Intern, Yammer team · Stage: Phone / technical screen
Microsoft Data Scientist internship phone screen for the SF Yammer team. The four questions focused on prior DS experience, explaining DS to a child or nontechnical audience, languages, and packages; the author described communication and logical expression as the emphasis.
Other reported topics or limits: Data-Science Background and Communication Screen.
106: Microsoft Research Systems Intern — phone or technical screen
Report date: 2020-03-28 · Role: Microsoft Research Systems Intern · Stage: Phone / technical screen · Interview date: 2019-12-16
MSR Systems internship pass report, published in 2020. The single Skype loop involved two researchers and was entirely about research, papers, and the candidate's next project steps; the author explicitly reports no coding.
Other reported topics or limits: Discuss Research, Papers, and Next Project Progress.
107: Microsoft Research + AI Research Intern — final interview loop
Report date: 2020-02-28 · Role: Microsoft Research + AI Research Intern · Stage: Final loop / onsite
Detailed Microsoft Research + AI internship video interview.
- Implement Softmax Cross-Entropy. Implement softmax cross-entropy with NumPy. The stated input is logits and the output is loss. The interviewer may also ask for a PyTorch implementation. Practice this problem
Other reported topics or limits: Candidate Paper and Research Deep Dive.
108: Data Scientist Intern, A/B testing team — final interview loop
Report date: 2020-02-21 · Role: Data Scientist Intern, A/B testing team · Stage: Final loop / onsite · Interview date: 2020-02-11
Microsoft Data Scientist internship onsite for an A/B-testing team. The four rounds covered behavioral and basic ML questions, an undisclosed A/B-testing case, an unnamed easy-to-medium LeetCode-style coding task, and probability/statistics including a vaguely recalled dice-expectation item.
Other reported topics or limits: Undisclosed A/B-Testing Case and Coding Question; Vague Dice-Expectation and Random-Number Questions; Behavioral and Basic ML Questions.
109: Data Science PhD Intern — final interview loop
Report date: 2020-02-13 · Role: Data Science PhD Intern · Stage: Final loop / onsite · Interview date: 2020-02-11
Detailed Microsoft Data Science PhD internship virtual onsite across Bing Ads and a market/Excel team.
- Number of Islands. The coding task was Number of Islands. Practice this problem
- Design Semantic Ad Retrieval for a Query. Given a query word such as 'shoes', retrieve relevant content from all ads. The discussion considered knowledge-graph word expansion, word embeddings, and learned similarity between the query and ad text or images.
- Design an Excel Next-Command Recommender. Recommend which Excel command a user is likely to use next based on actions in the current workbook/session. Available fields include time, user ID, Excel session ID, and command. Explain how to evaluate the recommendation system.
Other reported topics or limits: ML Concept and Resume Deep-Dive Questions.
110: Microsoft Research NLP Intern — phone or technical screen
Report date: 2020-02-03 · Role: Microsoft Research NLP Intern · Stage: Phone / technical screen
Detailed Microsoft Research NLP internship report.
- Return the Top K Search-Query Suggestions for a Prefix. A collection contains many search queries. Given the first several words of a string/query, return the top k likely suggested queries. The author identified a trie as the intended data structure. Practice this problem
- Implement a Direction-Specific Research Model. The interviewer asked whether the candidate knew a model related to the target internship direction and then asked them to implement it. The report did not preserve a reconstructable task contract.
- Implement a Sequence-to-Sequence Model. The second interviewer asked the candidate to implement seq2seq. The candidate wrote a short implementation and corrected one pointed-out mistake. The report did not preserve a reconstructable task contract.
Other reported topics or limits: Paper and Research Discussion.
111: Data & Applied Science Intern — phone or technical screen
Report date: 2020-01-30 · Role: Data & Applied Science Intern · Stage: Phone / technical screen · Interview date: 2020-01-22
Detailed Microsoft Data & Applied Science internship phone screen. The discussion focused on commonly used ML/statistical methods, one relevant project in depth, project challenges, explaining deep learning, and improving a favorite Microsoft product; no coding was asked.
Other reported topics or limits: ML Method, Project Deep-Dive, and Product-Improvement Screen.
112: Data Scientist PhD Intern, Bing team — final interview loop
Report date: 2020-01-28 · Role: Data Scientist PhD Intern, Bing team · Stage: Final loop / onsite · Interview date: 2020-01-16
Detailed Microsoft Data Scientist PhD internship process for a Bing/NLP-ML team.
- Knapsack Problem Variants. The interview included several versions of the knapsack problem, with dynamic programming as the focus.
- Largest K in an Array Using a Heap. Find the largest K values in an array using a heap. Practice this problem
Other reported topics or limits: Unspecified NLP List-Processing Task; ML and DL Concept-Comparison Questions; Recruiter Behavioral and Product Screen.
113: Data Analysis / Machine Learning Intern — final interview loop
Report date: 2020-01-26 · Role: Data Analysis / Machine Learning Intern · Stage: Final loop / onsite
Microsoft Japan summer internship onsite report for an input-method team seeking data-analysis help.
- Print Vertical Asterisk Bars. The input is a sequence of numbers representing bar heights. The requested output draws the bars with asterisks. For heights 3, 1, 2, the example uses three rows with bottom-aligned vertical bars. Practice this problem
- Word Search with Eight-Direction Adjacency. The input contains an N by N matrix filled with characters and a target string. The task is to determine whether the target string exists in the matrix. Horizontal, vertical, and diagonal moves are all valid, and a brute-force solution was acceptable. Practice this problem
Other reported topics or limits: Score Comparison and ML Discussion.
114: Data and Applied Scientist Intern — phone or technical screen
Report date: 2020-01-24 · Role: Data and Applied Scientist Intern · Stage: Phone / technical screen · Interview date: 2020-01-24
Detailed Microsoft Data and Applied Scientist internship recruiter screen. The interviewer asked the candidate to choose an ML/DL domain, then covered project challenge and validation, explaining deep learning to a child, language preferences, end-to-end ML workflow, motivation, and product improvement. No coding was asked.
Other reported topics or limits: Recruiter ML Background, Workflow, and Product Screen.
115: Data Scientist PhD Intern — phone or technical screen
Report date: 2020-01-23 · Role: Data Scientist PhD Intern · Stage: Phone / technical screen
Microsoft Data Scientist PhD internship phone screen. After a project and p-value discussion, the interviewer asked one fully described probability question about repeated rolls of two dice and which target sum appears first.
Other reported topics or limits: Probability That Sum Seven Appears Before Sum Eight; Project and P-Value Discussion.
116: Data and Applied Scientist Intern — phone or technical screen
Report date: 2020-01-22 · Role: Data and Applied Scientist Intern · Stage: Phone / technical screen
Microsoft Data and Applied Scientist internship first-round recruiter screen. The questions covered graduation timing, DS versus ML preference, product improvement, a challenging project and its data/language details, difficult teammate handling, prior ML tools and algorithms, and company knowledge.
Other reported topics or limits: Recruiter Background, Project, Product, and Teamwork Screen.
117: Data Scientist Intern — phone or technical screen
Report date: 2020-01-17 · Role: Data Scientist Intern · Stage: Phone / technical screen
Detailed Microsoft Data Scientist internship technical phone screen. The interviewer conducted a sustained linear-regression and estimation theory discussion covering coefficient interpretation, normalization, predictor choice, nonlinearity, interactions, signs, least squares, MLE, and their equivalence.
Other reported topics or limits: Linear Regression and Estimation-Theory Discussion; Graduation Date and Project Deep Dive.
118: Data & Applied Science PhD Intern — phone or technical screen
Report date: 2020-01-14 · Role: Data & Applied Science PhD Intern · Stage: Phone / technical screen · Interview date: 2020-01-14
Microsoft Data & Applied Science PhD internship technical phone screen. The discussion covered LSTM choice and metrics, gradient descent, poor validation performance despite strong training performance, and L1 versus L2 regularization.
Other reported topics or limits: LSTM, Gradient Descent, Generalization, and Regularization Discussion; Background, Experience, and Future-Project Discussion.
Patterns across the reports
- Three interviews back to back. Six current reports describe an initial screen followed by three consecutive interviews (006, 008, 009, 012, 017, 018).
- The precision-recall curve keeps coming back. It was named in a 2025 report (009) and in two older reports (056, 064).
- Research roles talk research. Research-track loops centered on the candidate's own work: three research talks in 015, research alignment and an ML design for the summer project with no coding test in 020, and LLM breadth plus one easy coding task in 019.
- Older loops were case-heavy. The History tables hold 26 named data-science and product cases, such as diagnosing churn (095) or measuring causal effects (094), against one in the current reports (008).
Common questions
How many rounds are in the Microsoft data science intern interview?
Usually an initial screen, then three interviews back to back (006, 008, 009, 012, 017, 018). Candidate 012 had a 30-minute screen and three 45-minute finals.
Does Microsoft give AI/ML interns an online assessment?
Two current reports had one: two one-hour questions plus SQL (008) and a HackerRank OA with a LeetCode task and a data-processing task (009). The others don't mention one.
Does Microsoft ask LeetCode questions for applied scientist interns?
Mostly yes, at easy to medium level: an easy pair-difference task (006), an easy Python string task (010), two-pointer and sliding-window array problems (014) and medium questions in each of 017's rounds.
What ML questions does Microsoft ask interns?
Fundamentals such as KNN and PCA (009), bias–variance, linear regression and overfitting (018), how ChatGPT is trained (012), and precision and recall in code (006, 009).
Do Microsoft research intern interviews include coding?
Not always: 020's research-team loop had no coding test, 015's final round was research talks, and 019 had one LeetCode Easy task next to LLM research questions.
Other records (not counted)
These are not counted in the 114: a loop relayed secondhand and two reports outside the AI/ML scope. Their questions are kept for practice.
Secondhand: a Bing Applied Scientist internship loop
This account was relayed by a friend rather than written by the candidate, so it is not counted as an independent experience. The reported final loop included:
- Spell correction with the nearest vocabulary word. Given a vocabulary and an input word, return the closest word when the input is absent. Practice this problem
- Prefix matching. Given a vocabulary and a prefix, return every matching word. Practice this problem
- Reservoir sampling. Keep K uniformly selected items from a stream. Practice this problem
- Distributed reservoir sampling. Extend the method across machines that do not know the global item count or complete reservoir state.
- Tree-node level or height checks. Determine whether selected nodes share the same level, then whether they have the same height. Practice this problem
- HTML element similarity. Design features for deciding whether two HTML elements are similar.
- Majority Element. Find the value that appears more than half the length of an array in O(n) time and O(1) extra space. Practice this problem
- Coin probability. Player A flips n+1 coins and player B flips n coins. Find the probability that A gets more heads. Practice this problem
- Windows menu-button experiment. Decide whether moving the Windows menu button from the left to the center is a good change. Design an A/B test and estimate the required sample size.
Other discussion covered logistic regression, modeling assumptions, projects, and behavioral questions.
Adjacent role: Explore Intern
The Microsoft Explore first round was a general undergraduate CS and behavioral phone screen. It did not contain AI/ML, data-science, applied-science, or research content, so it is kept here rather than counted in the main series.
Role not disclosed: a PhD internship report
One PhD internship report mentioned LeetCode 273 and behavioral questions, but it did not identify the team or establish an AI/ML-related role.
Keep practicing
- Practice from the current reports: Top K Frequent Elements (005); Return Every Pair with the Minimum Numeric Difference (006); Compute Dog Retrieval Precision and Recall (006); Design a Retrieval-Augmented Generation System (006); Implement K-Means from Scratch (007); Show an Organization Chart One Layer at a Time (008); Implement a Precision-Recall Curve (009); Find the Median from Two Arrays (018); Reverse Word Order In Place (021).
- Practice from older reports: every linked question in the History tables above.
- Microsoft banks: the coding bank, system design bank and behavioral bank (sign-in needed).
- Microsoft overview: the Microsoft interview page and Microsoft hiring insights.
- More reports: the Experience feed.
Sources: Anywhere useful interview info can be found.
Comments
0No comments yet — be the first to share what you know.