IMC Machine Learning Engineer interview questions

5 Machine Learning Engineers have shared notes from their IMC interviews, with the questions they were asked in each round. That is 25% of the 20 IMC interview notes on Screna AI.

5 Machine Learning Engineer interview notes

IMC·Machine Learning Engineer·Take-home Assignment

Jun 2026
IMC ML Engineer take-home: a fully open-ended coding task on a tabular dataset covering the whole pipeline from EDA through neural net training. No hand-holding, no starter code, just a dataset and a list of things they wanted to see.
  • Load a tabular dataset, summarize feature distributions, identify missing values and outliers, and visualize relationships between features and the target variable.
  • Implement a regularized linear regression model and evaluate its performance on a held-out test split.
  • Build and train a small feed-forward neural network on the same regression target, selecting an appropriate loss function, optimizer, and learning rate.
  • Detect overfitting from training and validation curves, then apply at least two techniques to address it and explain the trade-offs involved.

“Felt pretty comfortable here.”

View Post

IMC·Machine Learning Engineer·Technical Phone Screen

Apr 2026
IMC machine learning engineer interview that went into some surprisingly pure math territory. The probability question was manageable but the follow-up into linear algebra and portfolio construction felt like a quant interview snuck into an MLE loop.
  • You have a probability model with three states where the probabilities are q, q squared, and 1 minus q minus q squared. What is the valid range of q such that all three values are legitimate probabilities?
  • Given a payoff matrix A for three assets across those three states, express a portfolio constraint in matrix form such that the portfolio has non-negative payoff in every state.

“This one looked easy and I almost rushed it.”

View Post

IMC·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Interviewed for an MLE role at IMC and got hit with two classic probability problems straight out of the green book. Nothing flashy, but these things are deceptively hard to derive cleanly under pressure.
  • Gambler's Ruin: a player starts with $i, bets $1 per round on a fair coin flip, and stops when they either reach $N or go broke. What's the probability they eventually go broke? Then extend your answer to a biased coin where the win probability is not 0.5.
  • For a symmetric random walk on the integers starting at 0, derive the probability of ever returning to the origin, the expected time of that first return, and explain why the walk is recurrent in one and two dimensions but transient in three or more dimensions.

“I remembered the fair coin case pretty quickly, probability of ruin is just (N-i)/N, which feels intuitive in hindsight.”

View Post

IMC·Machine Learning Engineer·Technical Phone Screen

Apr 2026
IMC ML Engineer interview focused purely on conceptual fundamentals, two big topics: linear regression theory and transformer architecture. No coding, just talking through derivations and design decisions for what felt like a solid hour.
  • Derive the closed-form solution for ordinary least squares and walk through the assumptions the model relies on.
  • How do L1 and L2 regularization change the OLS objective, and what does that mean for the solution and the model's behavior?
  • Explain how self-attention works in a transformer, including the role of queries, keys, and values.
  • Why is the computational cost of self-attention quadratic in sequence length, and how does that compare to RNNs for long sequences?
  • What are positional encodings for, and what design choices exist around them?
  • Walk through the purpose of residual connections and layer normalization in a transformer block.
  • What are the main architectural variants of transformers and when would you use each?

“I got through the derivation fine, X-transpose-X inverse times X-transpose-y, but stumbled a bit listing all the assumptions cleanly.”

View Post

IMC·Machine Learning Engineer·Technical Phone Screen

Apr 2026
IMC put me through a market-making probability problem for an ML Engineer role. The core question was clean but the follow-ups got progressively harder and I'd be lying if I said I handled all of them gracefully.
  • You're a buyer in a market-making scenario. Each item is defective with probability 0.7 (worth 7000 to you) or genuine with probability 0.3 (worth 10000 to you). A defective seller only accepts bids of at least 3000, and a genuine seller only accepts bids of at least 7000. What single bid price maximizes your expected profit, given you can't observe item quality before bidding?
  • Now suppose you observe a partial signal about the item before bidding. Re-derive the optimal bid using conditional probability based on that signal.
  • Replace the discrete item values with values drawn from an exponential distribution and redo the expected profit analysis for your bidding strategy.

“The setup sounds manageable until you realize you have to compare two regimes, bid below 7000 and only defective sellers trade with you, or bid at least 7000 and both types trade.”

View Post