Stripe The styles of MLE OA and SWE OA are similar, but they are more focused on actual payment/risk control business scenarios. The amount of questions is usually 1 big question + multi-stage (3-5 parts), or 2 medium questions, and the time is 60-90 minutes. The difficulty of the questions is medium to above, focusing on data processing, feature engineering thinking, boundary processing and code engineering capabilities, rather than pure LeetCode algorithm.
Stripe MLE OA exclusive high-frequency real questions (Programhelp)
Topic 1: Identify traffic signs
Problem description
HackerCars is developing a self-driving car. As part of its autonomous sensing system, the company wants cars to automatically recognize speed limit signs on the road and adjust their speed accordingly.
Now given a set of data sets containing speed limit sign images and their corresponding labels, we need to use PyTorch Build a model to identify whether a speed limit sign is 30km/h, 70km/h or 120km/h.
Related documents
- Train.csv (training set)
- Test.csv (test set)
- The sample_submission.csv (sample submission file) task requires building a deep neural network to accurately recognize traffic speed limit signs. For each record in the test set (test.csv), predict the value of its label variable. Finally, a CSV file (submissions.csv) needs to be submitted, which should contain a table header, and each row corresponds to the prediction result of a test set record. The file format requires that the CSV file submitted contain exactly two columns:
- Path: image path
- Label: prediction label (0 corresponds to 30km/h, 1 corresponds to 70km/h, 2 corresponds to 120km/h). The evaluation index for evaluating the performance of the index model is accuracy (Accuracy). The calculation formula is: Accuracy = the total number of all predictions and the number of correct predictions
Topic 2: Query employee attendance
Problem description
Given a pandas data table containing employee attendance data of different sales branches, you need to find the employees who have visited their sales branches at least 3 times in a month.
Things to note
- A branch that an employee has visited once is regarded as its sales branch.
- An employee may visit the same branch multiple times on the same day, but it is only counted once.
- If there is no employee who meets the conditions, an empty DataFrame with a header needs to be returned. Data table structure The data table contains three columns:
- Emp_id: employee ID
- Branch_code: unique identifier for each branch
Topic 3: Identify traffic speed limit signs
Problem description
HackerCars is developing self-driving cars. As part of the vehicle's automatic perception system, the car needs to automatically recognize speed limit signs on the road and adjust its driving speed accordingly.
Given a data set containing speed limit sign images and corresponding labels, it is required to use PyTorch to build a model to identify speed limit signs as 30km/h,70km/h Or 120km/h Category three.
Provide documents
Train.csv:Training set fileTest.csv: test set fileSample_submission.csv:Submission file example
Mission requirements
Build a deep neural network to classify and recognize traffic speed limit signs.
For test set Test.csv For each record in , predict its Label The value of the variable. Finally submit a CSV file Submissions.csv, it is required to include a table header, and each row corresponds to a prediction result of the test set.
Submission file format
The file must contain exactly 2 columns:
Path:image pathLabel: Prediction label (0 represents 30km/h, 1 represents 70km/h, 2 represents 120km/h)
Evaluation indicators
Model performance is evaluated based on **Accuracy**, and the calculation formula is:
Accuracy=Total number of predicted samples Number of correctly predicted samples
Topic 4: Statistics of employee attendance at the store
Problem description
Given an employee attendance data set in pandas DataFrame format (recording employee visits to different sales branches), it is required to filter out employees who visited the sales branch at least 3 times in a month.
Things to note
- As long as an employee has visited a certain branch once, the branch belongs to his or her sales branch.
- Multiple visits by an employee to the same branch on the same day will only be counted as one visit.
- If there are no employees who meet the conditions, an empty DataFrame with a header needs to be returned.
Data structure
DataFrame contains 3 columns:
Emp_id: Employee IDBranch_code：The unique identifier of the branchVisit_date: Date of visit, the format isYYYY-MM-DD
Example
Sheet
| Emp_id | Branch_code | Visit_date |
|---|---|---|
| 1 | B01 | 2024-01-01 |
| 1 | B01 | 2024-01-03 |
| 1 | B01 | 2024-01-04 |
| 2 | B02 | 2024-01-02 |
| 2 | B02 | 2024-01-03 |
| 2 | B03 | 2024-01-01 |
Result analysis
- Employee 1: Visited the b01 branch 3 times within 1 month and met the conditions
- Employee 2: Visited branch b02 twice and branch b03 once, but did not meet the requirement of ≥3 times.
- The final result contains only records for employee 1
If you are preparing for Stripe MLE OA and need detailed solutions to the above questions + complete reference codes, or more 2026 latest real questions, please send me a private message to get them.
I can also provide:
- OA ghostwriting support(Ensure 100% test cases pass)
- MLE special mock interview (ML System Design + Coding)
- Full guidance from OA to VO
Send me a private message and tell me your time point and current level, and I will give you the most suitable preparation plan.
I wish everyone can speed up Stripe MLE OA and get offers as soon as possible!
