What I’m sharing this time is Point72 Overall experience and topic review of 2026 Summer OA. The overall process is very smooth, there is no need to turn on the camera, and the question setting is not a formality or free prostitution route, but is clearly centered around real work scenarios and ways of thinking. OA covers REST API data processing, Prompt design and basic algorithms. It not only examines engineering and analysis capabilities, but also focuses on the understanding of business and model applications. The overall difficulty is moderate. As long as the ideas are clear and the expression is reasonable, the passing rate is high and it has good reference value.
OA basic situation
- Project：Point72 2026 Summer Internship
- Form: Online OA, without turning on the camera during the whole process
- Duration: Relatively ample time and not stressful pace
- Question type：Data processing + Prompt design + basic algorithm
- Language: The coding language is relatively free (Python is the main one and more convenient)
- Experience: The question setting is biased toward real business scenarios, rather than just a question type OA
- Results feedback: The overall pass rate is friendly, and most of them will receive VO invitations after successful completion.
Point72 2026 Summer OA real question sharing
Q1. Football club revenue potential analysis (REST API + data analysis)
Obtain club information for all leagues through the REST API and calculate the potential revenue of each club: revenue = stadium capacity × (1 + 0.1 × number of top three league finishes + 0.5 × number of Champions League finishes + 0.05 × number of trophies)
Finally, the total revenue of all clubs is returned.
Problem-solving ideas
This question requires obtaining club information for all leagues via a REST API and calculating potential revenue based on stadium capacity, top three league finishes, Champions League finishes, and trophies. My idea is to first go through all the leagues, get the basic information of each club through the API, and then calculate the revenue of a single club based on the given formula. Once the calculation is complete, the revenue from all clubs is added up to get the total revenue. During the implementation process, I paid attention to dealing with data type issues. For example, the stadium capacity may be of string type and needs to be converted to an integer, and missing values are reasonably handled. The entire process is essentially data capture, calculation and aggregation, with an emphasis on clear logic and exception handling.
Q2. Design AI prompt words (Prompt Design)
Two cue words for coaches of 400-meter athletes:
- Analyze news articles, extract structured information and determine whether it is relevant to athletes.
- Collect the data needed to predict Olympic 400m championship times and national selection qualifications.
Problem-solving ideas
Tip 1: News article analysis
Output a structured template, for example:
{
"Title": "",
"Relevance": "High/Medium/Low",
"Do you mention me": "Yes/No",
"Quotation": "",
"Reason for recommendation": ""
}The model requires:
- Extract titles and quotes
- Judgment of relevance to athletes
- Suitable for news, blogs, forums and other article types
Tip Two: Data Collection and Forecasting
- Guide the model to identify key data sources:
- Historical results (Olympic Games, World Championships, Continental Championships)
- Athlete status (recent competition, training status)
- Home country selection rules (time, score requirements)
- Environmental factors (weather, venue)
- Output example:
{
"Historical achievements": [...],
"Athlete Status": {...},
"Selection Rules": {...},
"forecast data": {...}
}Q3. Dictionary minimization of strings (algorithm question)
Given lowercase string S And two subscript arrays Arr,Brr, you can select any multiple times I,exchange S Middle Arr[i] And Brr[i] Characters. Find the lexicographically smallest string obtained after any number of operations.
Problem-solving ideas
The core of this question is to find out which characters can be exchanged with each other to obtain the smallest string in lexicographic order. Each pair of interchangeable positions can be regarded as an edge in the graph, and each position of the string corresponds to a node in the graph. By constructing such a graph, one can use union-find or depth-first search to find all connected components, and the characters in each connected component can be interchanged with each other. After finding the connected components, extract the characters in each component and arrange them in ascending lexicographic order, and then put them back into the string according to their original positions. Since characters in the same connected component can be freely exchanged, the lexicographically smallest string under any exchange operation can be obtained after processing.
Reference code
Parent = list(range(n))
def find(x):
if parent[x] != x:
parent[x] = find(parent[x])
return parent[x]
def union(x, y):
parent[find(x)] = find(y)
Prepare efficiently and pass the interview easily
Overseas interviews are often stressful: communication and language barriers in different cultural backgrounds, as well as the need to master algorithms, system design and professional knowledge in a short period of time, making preparation both stressful and tedious. If you are worried that the interview will be rushed, or that preparing for the exam alone will be inefficient,ProgramHelp Can help you deal with it easily. We provide professional Interview assistance services , providing document answers, detailed problem-solving ideas and system analysis in real time, so that you don’t have to work alone, but you can also improve your interview pass rate in an orderly manner and quickly improve your self-confidence.
