Point72 Interview Questions — Page 3

Point72·Software Engineer·Take-home Assignment

May 2026

Point72 Data Engineer take-home involving two PySpark problems. The second one had five functions to implement and a few genuinely tricky spots if you're not careful about how Spark reads CSVs by default.

Set up a Spark session with the correct master and application name configuration.
Filter a medical bills table to keep only records that have a matching entry in an eligibility table.
Generate a full name column by combining first and last name from the eligibility table and return the result in the required schema.
Find the member who paid the highest amount, sorting by the payment column.
Calculate the total amount paid across all records and return it as an integer scalar.

“Straightforward.”

View Post

Point72·Software Engineer·Technical Phone Screen

May 2026

Point72 quant engineer interview that went pretty deep into linear regression, specifically around computational approaches when your dataset is massive. Not a brutal round but there was one moment where I definitely froze up before things clicked.

How do you estimate the coefficients in linear regression, and what approach would you use when the number of observations is much larger than the number of features?

“Started with gradient descent which they accepted, but then they pushed for a closed-form solution.”

View Post

Point72·Software Engineer·Technical Phone Screen

May 2026

Point72 data engineer interview with a Python coding problem that looked like a standard stock trading puzzle but had enough moving parts to trip you up if you weren't careful. The decorator requirement was the part I didn't see coming.

Given an array of daily stock prices, implement a function that finds the optimal set of non-overlapping buy/sell transactions to maximize total profit, and returns each transaction as a structured tuple with buy day, sell day, prices, and profit.
Wrap the transaction-finding function with a decorator that validates the input prices list against a set of constraints before the function runs.
Implement a report generator that takes the prices array and produces a human-readable string summarizing each transaction and the total profit.

“The core algorithm isn't too bad once you realize you can just capture every upward slope.”

View Post

Point72·AI Engineer·Technical Phone Screen

May 2026

Interviewed for an AI Engineer role at Point72 Asset Management. One of the questions was a LeetCode-style coding problem, not much else to go on from what I remember.

Solve a coding problem involving encoded numbers.

“Barely any detail survives from this one.”

View Post

Point72·Machine Learning Engineer·Hiring Manager Screen

Jun 2026

Behavioral screen for an ML Engineer role at Point72. Pretty standard opening round, three questions that all the big funds seem to run through before they let you near anything technical.

Walk me through your background and current role.
Why Point72 specifically? What draws you to this company over others?
Take me through the major projects on your resume. For each one, what was your role, what technical decisions did you own, and what was the actual impact?

“I prepped this but still rambled a bit.”

View Post

Point72·Data Scientist·Technical Phone Screen

Jun 2026

Two coding problems back to back for a Data Scientist role at Point72. The first was a k-th missing number problem and the second involved flagging redundant events inside a sliding time window. Neither felt purely DS-flavored, more like a straight SWE screen.

Given an unsorted array of distinct positive integers and an integer k, return the k-th smallest positive integer that does not appear in the array.
Given parallel arrays of operation names and timestamps, and a window size W, count how many events are redundant, where an event is redundant if the same operation has occurred at least twice within the preceding W seconds.

“My first instinct was to just iterate from 1 upward and count misses, which works fine for small inputs but falls apart when nums[i] can go up to a billion.”

View Post

Point72·Machine Learning Engineer·Technical Phone Screen

Jun 2026

Technical phone screen for an ML engineer role at Point72. One question, pretty conceptual, and the interviewer seemed happy with where the conversation went.

Given the same input, why might a decoder-based language model produce a different output each time it runs?

“Talked through sampling and temperature, how the model draws from a probability distribution rather than always picking the top token.”

View Post

Point72·Machine Learning Engineer·Technical Phone Screen

Jun 2026

Technical phone screen for an ML Engineer role at Point72. Just one question from what I can tell, but it went deep enough that I felt it.

What is the key difference between an encoder and a decoder in transformer architectures?

“The interviewer seemed pretty satisfied once I brought up the causal mask.”

View Post

Point72·Software Engineer·Technical Phone Screen

Jun 2026

Got a SQL problem for a Data Engineer role at Point72. One question, but it was a proper one, not just a basic join or group by situation.

Given a candidates table and a results table, write a SQL query that produces a per-candidate report showing their full name and top 3 states by vote count. Each place column should display as 'State (votes)'. If multiple states tie at the same rank, list them alphabetically separated by commas in that rank's column. Ranking should be based on distinct vote totals, not row position.

“The dense rank part is what got me initially.”

View Post

Point72·Data Scientist·Technical Phone Screen

Jun 2026

Deep technical interview at Point72 for a data scientist role. The whole thing was essentially one long project deep-dive with five distinct sub-questions, covering everything from feature engineering to SHAP to PCA math. Felt more like a PhD defense than a job interview.

Pick an ML project you built personally and walk through it end-to-end: problem statement, dataset, label definition, data quality issues, leakage risks, train/validation strategy, evaluation metrics, and model choices. Justify every decision.
How did you decide which features to include or exclude? Walk through your process for handling missing values, outliers, high-cardinality categoricals, correlated features, and time-based target leakage. How did you verify features stay stable over time?
For your chosen model, which hyperparameters mattered most, what search strategy did you use, how did you structure cross-validation for time-series data, and how did you avoid overfitting to the validation set when selecting the final model?
Write out the core PCA optimization objective and its solution. How does PCA connect to the covariance matrix and SVD? What do the principal components actually represent, how many do you keep, and when is PCA harmful in a supervised learning context?
Explain SHAP values and their connection to Shapley values from cooperative game theory. What properties make SHAP attractive? Interpret a summary plot, a dependence plot, and a force plot. What are at least three failure modes or pitfalls?

“This is the kind of question where picking the wrong project kills you.”

View Post