Applied Intuition Machine Learning Engineer interview questions

12 Machine Learning Engineers have shared notes from their Applied Intuition interviews, with the questions they were asked in each round. That is 19% of the 64 Applied Intuition interview notes on Screna AI.

12 Machine Learning Engineer interview notes

Applied Intuition·Machine Learning Engineer·Take-home Assignment

Jul 2026
Applied Intuition ML engineer interview centered on a take-home Colab notebook with a broken GPT-style model. The task was to find and fix three specific bugs, write tests, and explain the reasoning. Pretty hands-on, no fluff.
  • Given a minimal GPT-style language model in a Colab notebook, identify and fix three bugs: a broken attention mask, a training loop error, and missing positional encoding. Also write unit tests that would have caught each bug and explain your reasoning.

“This took me longer than I expected on the attention masking piece.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Jul 2026
Applied ML engineer interview with a pretty deep CUDA question about deterministic floating-point reductions. The whole thing felt more like a systems/numerics research chat than a standard coding screen, which I wasn't fully prepared for.
  • How would you design a deterministic parallel reduction in CUDA that guarantees a fixed combination order, given that floating-point addition is not perfectly associative?
  • How do you combine reduction results across CUDA blocks without using nondeterministic atomic operations?
  • What guarantees can and cannot be made about floating-point reductions, even with a deterministic ordering scheme?
  • What are the time and memory tradeoffs of a deterministic reduction approach compared to a standard parallel reduction?

“The interviewer nudged me toward thinking about prefix sums, which helped a lot.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Applied Intuition ML engineer interview, technical phone screen focused entirely on Transformer internals. One meaty question about positional encodings that took up basically the whole session.
  • Implement positional encodings for a Transformer language model. Walk through sinusoidal vs learned approaches, write PyTorch code to compute and add them to token embeddings, explain the math and tensor shapes, integrate into the full model, and describe what breaks if you skip positional info entirely and how you'd verify the fix.

“This was a lot to unpack in one question.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Applied Intuition ML engineer interview with a deep technical question on Transformer attention masking. Single question but it went pretty far into implementation details, edge cases, and debugging, more thorough than I expected for what felt like a phone screen.
  • Implement attention masking for an autoregressive Transformer that handles variable-length sequences with padding. Build both a causal mask and a padding mask, show how to combine them, and explain what bugs look like in practice and how you'd test for them.

“This one went deeper than I was ready for.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Interviewed for an ML Engineer role at Applied Intuition. The questions leaned heavily on research background and future direction, which felt more Research Scientist than MLE to me, but I went with it.
  • Walk me through your most relevant past research experience. What problems did you work on, what methods did you use, and what measurable impact came out of it?
  • What research areas are you most excited to explore next, and what's drawing you toward them?
  • How do your research interests align with what this team is currently working on?

“This is the kind of question that sounds easy until you're actually in it and realize your 'measurable impact' is a 2% improvement on a benchmark nobody outside your lab cares about.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Interviewed for an ML Engineer role at Applied and got a data structures implementation question, which surprised me a bit. Not what I expected going in, but it was a fair problem if you've thought about circular buffers before.
  • Design and implement a fixed-capacity double-ended queue (deque) from scratch. It should support insertions and deletions from both ends, plus peek and status checks, all in O(1) time. Handle the case where the deque is full or empty gracefully, and make sure wraparound works correctly.

“My first instinct was to use a plain array with two pointers, which is the right call, but I fumbled the wraparound logic for a bit.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Applied gave me a technical screen for an ML Engineer role that turned into a pretty involved coding problem about schema validation. No fluff, just code. The problem was well-scoped but had enough edge cases to keep you honest.
  • Given a set of custom descriptor objects that define an expected schema, implement a validator that checks whether an input configuration matches the descriptor. The validator should handle primitives, nested objects, and lists, and return a list of errors that include field paths and type information.

“This one took me a minute to fully parse.”

View Post

Applied Intuition·Machine Learning Engineer·Onsite - System Design / Architecture

May 2026
Applied Intuition ML engineer interview, system design round focused entirely on KV caching for transformer inference. Pretty deep technically, they wanted actual tensor shapes and API signatures, not just hand-waving about attention.
  • Design and implement a key-value caching system for autoregressive inference in a Transformer decoder. Include tensor shapes, a per-layer cache API that works across decoding steps and batched inputs, memory limit handling, and support for both greedy and beam search.
  • Walk through the complexity analysis and expected speedup of KV caching compared to full recomputation at each decoding step.
  • How would you test correctness of the KV cache implementation, particularly for edge cases like an end-of-sequence token appearing in the middle of a batch?

“This one took me a minute to scope properly.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Applied Intuition MLE interview threw a curveball with a CUDA-specific question I had zero prep for. Short session but it left me thinking I need to go deeper on GPU parallelism fundamentals.
  • Describe a CUDA algorithm that guarantees a specific reduction order, similar in concept to prefix sum.

“Had no idea.”

View Post

Applied Intuition·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Coding round at Applied Intuition for an MLE role. One problem, but it was a deceptively tricky config validation thing that required a lot of back-and-forth with the interviewer just to understand the scope.
  • Implement a config validator that takes a set of descriptors (potentially nested, e.g. a PersonDescriptor with name, email, and an AddressDescriptor) and validates a JSON input against them.

“The problem sounds manageable until you realize all the descriptor classes are custom and you have to figure out the contract yourself.”

View Post