← Back to Directory

Two Sigma interview notes, page 1 of 7

Large Enterprises

Most of these Two Sigma write-ups come from Data Scientists (31), then Software Engineers (30) and Machine Learning Engineers (1). The most reported rounds are the technical phone screen (53), the onsite coding round (5), the online assessment (2) and the hiring manager screen (1).

62 interview notes · updated Jul 2026

Two Sigma·Data Scientist·Technical Phone Screen

Sep 2025
Two Sigma data scientist round, heavy on vectorized pandas and SQL. The whole thing was basically one long coding problem about NYC taxi data with a few sub-tasks layered inside it. More math-y than I expected for a DS role.
  • Given trips and zones datasets, join them on pickup zone, then compute median trip speed (mph) per borough and hour of day. Filter to trips where duration is between 1 and 120 minutes and speed is between 1 and 80 mph. Return the top 3 (borough, hour) pairs by median speed, breaking ties by borough then hour alphabetically.
  • For Manhattan pickups between midnight and 5am, find the 3 taxi IDs with the highest 95th percentile trip duration in minutes. Break ties by taxi_id ascending. Define clearly how you compute the 95th percentile.
  • Write the full pandas implementation for the above analyses in O(n log n) or better. Requirements: no row-level Python loops, parsed datetime dtypes, a single join performed once, categorical dtype for borough, and indexing on pickup_ts for time filtering.
  • Justify two specific memory or performance optimizations in your implementation, such as downcasting numeric columns or doing groupby-agg in a single pass instead of multiple passes.

“The filtering part seemed easy but I almost forgot to convert duration to hours before computing speed.”

View Post

Two Sigma·Software Engineer·Technical Phone Screen

Apr 2026
Two Sigma SWE coding round. One problem, 30 minutes, and I did not finish. The question looked like a data structures exercise on the surface but the implementation details pile up fast and I ran out of time.
  • Design and implement a simplified order matching engine that processes buy and sell orders, matching them by price-time priority and outputting fill results or the current order book state.

“I knew the general idea going in: max heap for buys, min heap for sells, match when best bid crosses best ask.”

View Post

Two Sigma·Software Engineer·Technical Phone Screen

Apr 2026
Two Sigma coding round, one algorithmic problem the whole session. The problem itself was well-designed and I actually enjoyed it once I figured out the core insight, though getting there took longer than I'd like to admit.
  • You start with w capital and have n projects, each with a required minimum capital and a profit value. You can complete at most k projects, picking any available one whose capital requirement you currently meet and collecting its profit. What's the maximum capital you can end up with?

“Took me a bit to stop thinking about this as a DP problem.”

View Post

Two Sigma·Data Scientist·Technical Phone Screen

Apr 2026
Two Sigma DS case study round centered almost entirely on a house-price prediction problem. The emphasis was heavily on data cleaning and feature work rather than model sophistication, which I didn't fully anticipate going in.
  • Given a house-price prediction problem, how do you decide which features are actually usable versus which ones you'd drop before even touching the model?
  • Walk through your data cleaning process for this dataset: missing values, outliers, skewed targets, encoding categorical variables.
  • What feature engineering would you do on top of the raw features for a house-price model?
  • The interviewer is only comfortable with linear regression and won't approve anything more complex. How do you still build a defensible model under that constraint?
  • How would you evaluate the model, and what metrics would you report to someone who isn't technical?

“I went straight to numeric vs categorical as if that was the main distinction, and the interviewer kept pushing until I got to leakage.”

View Post

Two Sigma·Software Engineer·Technical Phone Screen

Apr 2026
Two Sigma quant engineer interview with a probability/statistics theory question that had a real proof component to it. Not the usual plug-and-chug stuff, they actually wanted rigorous justification.
  • Let X and Y be random variables and f(X, Y) a measurable function. First, is it always true that the minimum over y of E_X[f(X, y)] is less than or equal to E_{X,Y}[f(X, Y)]? Prove it or give a counterexample. Second, does this inequality hold with equality conditions when X and Y are independent? Justify rigorously.

“The first part clicked pretty fast for me.”

View Post

Two Sigma·Data Scientist·Technical Phone Screen

Apr 2026
Two Sigma data scientist case interview focused entirely on a housing price prediction problem. It was a long, discussion-driven session where every design choice got challenged with follow-ups. Not a coding round at all, more like a sustained technical debate.
  • You have a multi-year dataset of residential real estate transactions with property attributes, location info, and sale dates. Build a model to predict housing prices, and walk through your end-to-end approach.
  • How do you precisely define the prediction target, and what features would you propose from raw transaction records to build a supervised learning dataset?
  • Why would a non-linear model outperform linear regression on housing data specifically? And how would you set up your evaluation metrics and validation scheme?
  • How do you handle missing values in the dataset, approach feature selection, and guard against data leakage?
  • After deployment, how would you detect input drift and relationship drift in a housing price model? And how would your approach handle a shock like the COVID-19 pandemic?
  • Your non-linear model only marginally outperforms the linear baseline. How do you decide whether that complexity is worth it in production?
  • The model is systematically underpricing homes in one specific region. Walk through your debugging process.
  • The business wants calibrated prediction intervals, not just point estimates. How do you produce and validate them?
  • When would you argue for building separate per-region models instead of a single national model, and what are the costs of that approach?

“This is the whole interview basically.”

View Post

Two Sigma·Data Scientist·Technical Phone Screen

Apr 2026
Two Sigma data scientist interview, got hit with a grid/island problem that felt more like a competitive programming warmup than anything data-sciency. Not a bad experience, just unexpected territory.
  • Given a binary grid of 0s and 1s, find the number of distinct island shapes, where two islands count as the same shape if one can be translated (no rotation or reflection) to exactly overlap the other.

“My first instinct was to just count connected components, which is obviously wrong.”

View Post

Two Sigma·Data Scientist·Online Assessment (OA)

Apr 2026
Two Sigma data scientist interview with a statistics and regression problem that looked straightforward on paper but had enough moving parts to trip you up if you weren't careful about the implementation details.
  • Given a list of unsorted day-temperature pairs, compute the median temperature, sample variance, OLS regression slope and intercept, and a predicted temperature for a given query day index.

“The median part is fine, just sort the temps and handle even/odd.”

View Post

Two Sigma·Software Engineer·Onsite - Coding / Algorithms

Apr 2026
Two Sigma coding round, one meaty simulation problem about IPO order matching. The core idea isn't that complex but the tie-breaking rule inside same-price groups is where they really want to see if you're paying attention.
  • Given a list of buy orders (each with a price, timestamp, and quantity) and a fixed share supply, simulate IPO order matching: fill higher bids first, and within the same price tier, distribute shares round-robin across orders sorted by timestamp. Output the final allocation per order.

“I got the high-level structure pretty fast.”

View Post

Two Sigma·Data Scientist·Technical Phone Screen

Apr 2026
Technical screen at Two Sigma for a data scientist role. The whole session was basically one extended stats problem about OLS regression, which sounds simple but went pretty deep pretty fast.
  • You run two OLS regressions on the same dataset: y on x, and then x on y. Derive both slope estimates using sample moments. What is the product of the two slopes, and why is the second slope generally not just the reciprocal of the first?
  • Under what condition on the data are the two OLS slopes exactly equal? How would you transform the variables to guarantee this, and what is the common slope value in that case?
  • If both x and y are noisy measurements of the same underlying true variable, is either OLS slope an unbiased estimator of the true relationship? What alternatives exist and what do they assume?
  • Geometrically, how do the two fitted lines relate to each other in the x-y plane? When do they coincide?
  • What happens to each of the two slopes if you add independent noise to x only, leaving y unchanged?

“I knew the formula for the slope off the top of my head but fumbled explaining why the reciprocal relationship breaks down.”

View Post