Citadel Interview Questions — Page 13

Citadel·Data Scientist·Technical Phone Screen

Jun 2026

Citadel data scientist interview, technical phone screen that went deep into probability and linear algebra pretty fast. The equicorrelation matrix question was the main event and it had a generalization attached that I did not see coming.

Given three zero-mean, unit-variance random variables where every pairwise correlation equals ρ, find the tightest lower bound on ρ that keeps the 3x3 correlation matrix positive semidefinite. Use eigenvalue analysis to show your work, then generalize the result to n variables.

“The 3-variable case I could handle.”

View Post

Citadel·Data Scientist·Technical Phone Screen

Jul 2026

Citadel data scientist interview with a meaty ML system design question about housing market prediction. One question, lots of surface area, and they clearly wanted to see if you could hold the whole thing together without falling apart on any single piece.

You have monthly housing market time series data across regions (price indices, mortgage rates, inventory, days-on-market, macro indicators). Design a classification system to predict whether a buyer should purchase now or wait up to k months. Walk through label definition, preprocessing, feature engineering, train/val splits, model choices, evaluation, leakage prevention, concept drift handling, and how you'd surface results to users.

“This question has so many sub-parts that I spent the first two minutes just trying to figure out where to start.”

View Post

Citadel·Data Scientist·Technical Phone Screen

Jul 2026

Citadel data science interview that was basically a full ML system design session. They wanted you to walk through two complete workflows from raw data to deployed model, regression and classification, and actually justify every decision along the way. Dense and technical, felt more like a take-home prompt than a phone screen.

Design and implement two end-to-end ML workflows on tabular data: one regression task predicting a continuous target and one classification task predicting a binary or multiclass label. For each, walk through data cleaning, train/val/test splitting, model selection, evaluation, tuning, reproducibility, and interpretability.

“This is a beast of a question and I did not pace myself well.”

View Post

Citadel·Data Scientist·Technical Phone Screen

Jul 2026

Citadel data science interview with some heavy probability/stats theory. The question was essentially a two-part derivation involving change of variables and the logistic transform, which felt more like a math exam than a job interview.

Given a random variable X with density f_X on (0,1) and the logistic function g(u) = 1/(1 + e^{-u}), derive the general density formula for U = g^{-1}(X) using a change of variables. Then, for X ~ Uniform(0,1), compute the distribution, mean, and variance of U = log(X/(1-X)), showing the Jacobian steps explicitly.

“This wrecked me a little.”

View Post

Citadel·Software Engineer·Technical Phone Screen

Jul 2026

Citadel software engineer interview with a concurrency-focused coding question. Pretty technical, they wanted you to actually think through the tradeoffs rather than just spit out a solution.

Make a counter thread-safe. Then walk through and compare the different approaches you could use to achieve thread safety.

“The first part is easy enough, locks, atomics, whatever.”

View Post

Citadel·Software Engineer·Technical Phone Screen

Jul 2026

Citadel software engineer interview that went deep on JavaScript internals and real-time systems. The questions felt less like trivia and more like they wanted to see if you actually understood the mechanics, not just the syntax.

What are JavaScript's data types, and which values evaluate to false in a boolean context?
For Promises, what do then, catch, and finally handlers return? What happens to the returned Promise if an error is thrown inside the executor or inside any handler, including inside an async function?
Compare plain objects and Map in JavaScript. How do they differ in terms of key types, iteration order, prototype properties, and performance?
What is a WeakMap, and how does it differ from Map in terms of what keys are allowed and how garbage collection works?
For WebSockets, does a single connection guarantee in-order and reliable message delivery? When might ordering break down, and how would you design a system to handle ordering and idempotency if needed?

“Felt easy until I second-guessed myself on empty string vs zero vs null vs undefined.”

View Post

Citadel·Software Engineer·Technical Phone Screen

Jul 2026

Citadel quant researcher interview, got hit with a classic combinatorics/CS problem that I thought I knew but fumbled the complexity analysis on. The follow-up about scaling to huge n was the part that really separated candidates I think.

Implement a simulation of the Josephus problem: n people in a circle, eliminate every k-th person starting from person 1, return the surviving position.
Derive and implement the O(n) recurrence solution for the Josephus problem: J(1,k)=0, J(n,k)=(J(n-1,k)+k) mod n, iteratively with O(1) space.
When would you prefer the simulation approach versus the recurrence-based approach, and how would you handle the case where n is on the order of 10^9?

“I went straight to a list-based simulation and it worked fine.”

View Post

Citadel·Data Scientist·Technical Phone Screen

Jul 2026

Citadel data science interview that went deep on statistical modeling fundamentals, specifically around regression diagnostics and tree-based methods. The house price design question was the centerpiece and took up most of the time. Technically demanding throughout.

In linear regression, how do you detect and handle outliers and influential points? Walk through Cook's distance and high-leverage diagnostics.
For Random Forests, how do you approach pruning individual trees and computing variable importance?
Design a modeling framework to predict house prices in a city. What features would you include and why?
When the number of predictors is very large, how do you compute or update linear regression coefficients in mini-batches without loading all the data into memory at once?

“I knew Cook's distance conceptually but fumbled when they pushed on the actual threshold intuition.”

View Post

Citadel·Data Scientist·Technical Phone Screen

Jul 2026

Citadel Data Scientist intro screen, basically a structured self-pitch covering background, projects, and fit. Pretty standard opener but the scope of what they wanted covered in one answer was a lot.

Walk us through your background, your most relevant ML projects and their impact, your key strengths and areas you're still developing, how you work with others, why you want this role and company, and give an example of a hard ML problem you tackled and what you took away from it.

“This is a lot to pack into one answer and I underestimated how much they actually wanted.”

View Post

Citadel·Software Engineer·Technical Phone Screen

Jul 2026

Citadel SWE interview with two algorithm-heavy problems, both requiring you to think carefully about string structure and cost optimization under constraints. The problems were harder than I expected and I left unsure how clean my solutions actually were.

Given a lowercase string and an integer k, find the minimum number of character changes needed to make the string both a palindrome and periodic with period k (i.e., every character equals the one k positions ahead of it).
Given n images each active over a date range with an associated filter cost, and a fixed discount price you can apply on any day to replace that day's total cost, find the minimum total processing cost across all days modulo 1,000,000,007.

“This one took me a while to untangle.”

View Post