Meta Applied AI Engineer interview questions

Attention is permutation-invariant, so without position information a transformer cannot tell word order. The signal is knowing why, and why the field moved from absolute sinusoidal encodings to relative ones like RoPE and ALiBi that extrapolate to longer contexts.

Embeddings are the substrate of modern NLP and retrieval. This checks whether you understand the distributional idea behind them, not just 'words become vectors.' The signal is how training on context produces geometry that encodes meaning.

Offline metrics gate releases; online metrics tell you what actually happened with users. The signal is knowing why offline wins routinely fail to hold online, and how the two feed each other.

Cross-attention is how a model attends from one sequence into another, the mechanism behind encoder-decoder models and multimodal conditioning. The signal is knowing exactly where the queries vs keys/values come from.

Full attention is O(n^2), so long context needs cheaper attention. The signal is naming the families (sparse, sliding-window, linear), what each gives up, and why FlashAttention is not one of them.

Modern LLMs dropped learned absolute positions for RoPE and ALiBi to handle long context. The signal is knowing both encode RELATIVE position, and exactly how that enables length extrapolation past the trained window.

MQA and GQA shrink the KV cache, the thing that bottlenecks LLM serving. The signal is knowing they share key/value heads across query heads to cut memory and bandwidth, with GQA as the quality-preserving middle ground. Here is the answer.

Scaling laws explain why bigger models trained on more data predictably get better, and Chinchilla changed how we allocate compute. The signal is the model-size-vs-data tradeoff and that many models were under-trained. Here is the answer.

How you decode from the model's probabilities shapes the output as much as the model. The signal is knowing the deterministic vs sampling methods and when each fits (factual vs creative). Here is the answer.

BERT and GPT split on one decision made before a single weight is trained: what does each token get to see? That choice cascades into attention, use case, and why one family now dominates. Here is the answer.

Multilingual models are uneven across languages, and most candidates only name the obvious cause. The differentiator is the tokenizer: the same sentence can cost a non-English user several times the tokens. Here is the answer.

A transformer emits one vector per token, but retrieval needs one vector per text. The signal is knowing the pooling options and the catch that almost everyone misses: the model has to be trained for whichever one you pick.

Both are conditional compute, but one scales parameters and the other scales depth. The signal is knowing which axis each routes along and why that changes the FLOP story. Here is the sharp version.

RAG and fine-tuning get framed as either/or. RAFT trains the model to be good at RAG itself. The tell is what goes in the training set: relevant docs mixed with distractors. Here is the answer.

Distillation buys you most of a frontier model's quality at a fraction of the serving cost. The signal is naming the three LLM-specific variants and knowing which one works against a closed API. Here is the answer.

A few thousand carefully chosen examples can beat a million scraped ones. The signal is knowing which filters matter, how you measure example quality, and why diversity beats raw volume.

GRPO drops the value model and estimates advantage from a group of samples per prompt. The signal is the exact mechanics, why the group baseline works, and the normalization traps that quietly bias training.

Replace a learned reward model with a checker that returns right-or-wrong, and reward hacking mostly disappears. The signal is why verifiable rewards beat learned ones for math and code, and where they break.

Distillation is not one method. The signal is knowing when to match logits versus train on generated text, why on-policy distillation beats off-policy, and how reasoning models are distilled.

The ratio of web, code, books, and math in pretraining quietly decides downstream skills. The signal is knowing how mixtures are chosen, why upsampling helps, and how methods like DoReMi automate it.

Continued pretraining injects domain knowledge that fine-tuning cannot, but naively it wrecks general ability. The signal is the replay ratio, learning-rate rewarming, and how you measure forgetting.

If a benchmark leaked into pretraining, the score measures memorization, not ability. The signal is the detection methods (n-gram overlap, canaries, perturbation tests) and why fresh held-out evals matter.

Drop the oldest tokens to bound the KV cache and the model collapses. Keep just the first few tokens and it generates indefinitely. The signal is understanding why those first tokens act as an attention sink.

A model trained at 4K tokens can be stretched to 128K without full retraining. The signal is knowing why naive extrapolation fails and how PI, NTK-aware scaling, and YaRN each rescale RoPE frequencies differently.

Speculative decoding speeds up generation without changing the output distribution. The signal is knowing how draft-model, self-drafting (Medusa/EAGLE), and lookahead approaches differ in where the guesses come from.

FP8 is the format behind modern H100-class training and serving. The signal is knowing the two FP8 variants, why a floating format beats INT8 for dynamic range, and where scaling still matters.

VLMs like GPT-4V and LLaVA bolt a vision encoder onto a language model through a projector. The signal is the image-tokens-as-prefix design, the alignment training, and the resolution and hallucination failure modes.

Speech LLMs either transcribe to text or model audio directly as discrete tokens. The signal is the semantic-vs-acoustic token split and why end-to-end audio models beat ASR-plus-LLM pipelines on latency and prosody.

A deceptively simple ML-coding ask. Anyone can write exp/sum; the signal is the max-subtraction trick and the log-sum-exp form that keep it from overflowing. Here is the stable implementation and why the naive one breaks.

The most-asked design-coding question, and a frequent warm-up at the labs before the ML follow-ups. The signal is the hashmap-plus-doubly-linked-list for true O(1), then handling the TTL and concurrency follow-ups cleanly. Here is that build.

A classic that rewards the two-heap insight. A sorted list gives O(n) inserts; two balanced heaps give O(log n) insert and O(1) median. Here is the implementation and the rebalancing detail people get wrong.

The from-scratch implementation that frontier labs and NVIDIA actually ask. Writing it proves you understand shapes, the scale, the causal mask, and how heads split and recombine. Here is a correct, readable implementation and the follow-ups.

Beam search is the decoding algorithm everyone references and few can implement correctly. The signal is keeping k hypotheses by cumulative log-prob, summing logs (not multiplying probs), and handling completed sequences and length. Here is a correct implementation and the tradeoffs.

A from-scratch ML-coding staple that checks whether you actually know the math you use. The signal is the clean gradient (it simplifies to Xᵀ(ŷ - y)/n), numerical stability, and vectorization. Here is the implementation and the details interviewers push on.

Autocomplete is the canonical trie question, and it tests whether you reach for the right structure instead of scanning a word list. The signal is O(prefix) lookup, the DFS to collect completions, and the follow-ups (ranking, memory). Here is the build.

A classic that tests whether you reach past sorting for the right structure. The signal is the heap solution (O(n log k)), the bucket-sort O(n) trick, and how it changes for an unbounded stream. Here is the answer.

Sliding window is one of the highest-yield coding patterns, and this is its canonical problem. The real signal is recognizing when a window collapses an O(n squared) scan into one linear pass, and maintaining the invariant cleanly.

Graph traversal underpins a huge class of interview problems. The signal is knowing the BFS-vs-DFS tradeoff (shortest path vs memory shape) and applying it cleanly, then handling the directed-vs-undirected cycle gotcha most candidates miss.

DP feels like pattern-matching magic until you have a method. The signal is a repeatable drill: spot overlapping subproblems and optimal substructure, define the state and recurrence, then memoize or tabulate. Here is that method on a worked example.

A common ML-coding check that verifies you actually understand the metrics you quote. The signal is getting TP/FP/FN/TN right and knowing macro vs micro averaging for multiclass, not reaching for sklearn.

Data loading is where training pipelines silently bottleneck. The signal is a correct shuffle-then-batch iterator plus knowing why prefetching and parallel loading keep the GPU fed.

Interval merging is a classic that tests the sort-then-sweep insight. The signal is sorting by start, then merging in one pass, O(n log n). Here is the pattern and the family of problems it unlocks.

k-NN is simple to write and a good test of vectorization and the lazy-learner tradeoff. The signal is a clean vectorized distance computation and knowing why naive prediction is O(n) and how to speed it. Here is the implementation.

Implementing conv2d proves you understand what a CNN layer actually computes, not just that you can name it. The signal is correct output-shape math, clean stride/padding handling, and knowing the im2col trick frameworks really use.

LayerNorm sits in every transformer, and implementing it checks that you know what it normalizes over and why. The signal is normalizing across the feature dimension per token, the learnable scale/shift, and the RMSNorm simplification modern LLMs adopted.

Binary search is easy to describe and famously easy to get wrong (off-by-one, infinite loops). The signal is a correct template plus the variants (boundaries, rotated arrays) that come up constantly. Here is the answer.

Backtracking is the pattern behind permutations, subsets, N-queens, and constraint problems. The signal is the choose/explore/un-choose template plus pruning, and knowing exactly which knob separates permutations from combinations.

A warm-up that screens for the stack instinct. The signal is recognizing last-opened-first-closed maps to a stack, and handling the edge cases (leftover opens, early close). Here is the pattern.

Merging k sorted sources is the canonical min-heap problem and a real data-engineering pattern (merging sorted shards/streams). The signal is the heap of k heads giving O(N log k). Here is the answer.

A classic that rewards the two-pointer insight over brute force. The signal is realizing water at each position is bounded by the min of the max walls on each side, then computing it in one pass. Here is the answer.

Implementing BatchNorm checks that you know the train/inference difference, the part everyone forgets. The signal is normalizing over the batch with running stats maintained for inference. Here is the implementation.

The canonical warm-up that screens for the hash-map instinct: trade space for time to turn O(n squared) into O(n). The signal is the one-pass complement lookup. Here is the pattern.

Topological sort orders a DAG so dependencies come first, the backbone of build systems, schedulers, and ML/data pipeline DAGs. The signal is Kahn's algorithm (or DFS) plus cycle detection. Here is the answer.

Union-Find answers 'are these in the same group?' near-instantly and powers connected-components, cycle detection, and clustering. The signal is path compression plus union by rank for near-O(1) operations. Here is the answer.

A grid-traversal classic that tests BFS/DFS flood fill and the visited-marking discipline. The signal is treating the grid as a graph and counting connected components. Here is the answer.

Edit distance is the canonical 2D string DP, behind spell-check, diff, and fuzzy matching. The signal is defining the subproblem and the insert/delete/replace recurrence. Here is the answer.

Sampling uniformly from a stream you cannot fit in memory (or whose length you never learn) is a real data-engineering problem. The signal is the replace-with-decreasing-probability trick and a proof that telescopes. Here is the answer.

Reversing a linked list is the canonical pointer-handling screen. The signal is the three-pointer walk done without ever losing the rest of the list, plus knowing why the iterative version beats recursion in production. Here is the answer.

Encoding a tree into a string and rebuilding it exactly hinges on one decision most candidates skip. Miss it and the structure becomes ambiguous. Here is the answer and the interview-grade reasoning.

LCA is the recursion classic that tests how you propagate information up from subtrees in a single pass. The signal is the post-order 'found in left, found in right' logic. Here is the answer plus the BST shortcut.

Maximum subarray is the gateway 1D DP problem, and Kadane's is the O(n) one-liner. The signal is the 'extend or restart' decision at each element and the all-negative edge case most people botch. Here is the answer.

Word Break is the classic 1D string DP that trips people who reach for greedy or naive recursion. The signal is the dp[i] = 'is the prefix of length i segmentable' recurrence, plus why a locally valid split can doom the rest. Here is the answer.

Rotating a matrix in place tests index manipulation and the transpose-then-reverse trick. The signal is decomposing the rotation into two simple passes rather than juggling four-way swaps. Here is the answer.

Bit-manipulation problems screen for comfort with binary and the O(1)-space tricks most candidates fumble. The tell is whether you reach for XOR's self-canceling property and the n & (n-1) idiom on instinct. Here is the toolkit.

K-th largest has three textbook solutions, and the signal is knowing Quickselect's average O(n) beats sorting's O(n log n), why its worst case is O(n²), and when a heap is actually the better call. Here is the answer.

TF-IDF is the classic text-vectorization scheme and a fast check of whether you actually understand term-frequency times inverse-document-frequency, or just recite it. The signal is explaining why IDF kills common words. Here is the implementation.

The DP that exposes why greedy quietly breaks on non-standard denominations, and how one bottom-up table fixes it. The signal interviewers want is the dp[amount] recurrence and the unbounded-reuse insight.

LCS is the template 2D-sequence DP behind diff tools and bioinformatics. The signal is reconstructing the match/mismatch recurrence cold and not confusing subsequence (gaps allowed) with substring (contiguous). Here is the answer.

Gas Station rewards spotting a greedy invariant that turns O(n²) into O(n). Two facts carry the whole solution: total feasibility, and that a failed prefix lets you skip every start inside it.

Jump Game rewards spotting that a greedy reachability scan beats DP. The signal is tracking the farthest reachable index in one pass, and detecting the exact moment you fall behind it.

LIS has an obvious O(n²) DP and a clever O(n log n) solution that surprises people. The signal is the patience-sorting/binary-search approach maintaining 'tails'. Here is both.

Minimum Window Substring is the hard sliding-window problem that tests expand-and-contract with a character-count map. The signal is the grow-to-valid then shrink-to-minimal pattern, checked in O(1) per step.

House Robber is the cleanest choose-with-a-constraint 1D DP, and it collapses to O(1) space. The signal is the take-this-or-skip recurrence and the two-variable rolling optimization.

Group Anagrams tests the 'compute a canonical key and bucket by it' pattern. The signal is choosing a key that anagrams share without comparing every pair. Here is the answer.

Building a PR/ROC curve from scratch proves you understand thresholds and the precision/recall tradeoff. The real signal is sweeping the threshold in one sorted pass and integrating the area. Here is the implementation.

This problem bans the obvious division trick, forcing the prefix/suffix-product insight. The signal is computing left and right products in two passes with O(1) extra space. Here is the answer.

Sorting an array of three values in one pass tests the three-pointer partition (Dutch National Flag). The signal is the low/mid/high invariant and knowing when not to advance mid. Here is the answer.

Spiral traversal tests careful boundary management more than algorithmic insight. The signal is shrinking four boundaries layer by layer without double-visiting a row or column. Here is the answer.

The canonical grid DP where each cell depends on three neighbors at once. The signal interviewers want is the min-of-three recurrence, and the reason it has to be min. Here is the answer.

This famously-hard problem wants better than the O(m+n) merge: an O(log) binary search on the partition. The signal is binary-searching the split point so the left halves stay below the right halves. Here is the answer.

Decode Ways is a 1D DP whose difficulty lives entirely in the edge cases: zeros and the valid 1-26 range. The signal is the take-one-digit-or-two recurrence plus disciplined validity checks. Here is the answer.

This classic has a clean O(n^2) expand-around-center solution that beats the naive O(n^3). The signal is expanding from every center and handling odd and even lengths separately. Here is the answer.

Implementing Adam proves you understand what's happening under .step(): momentum plus per-parameter adaptive rates with bias correction. The signal is the two moments and the bias-correction step.

Finding the max of every window of size k naively is O(n*k). The trick that drops it to O(n) is a monotonic deque of indices, and the reason it stays linear surprises most candidates.

The minimum number of meeting rooms equals the peak number of meetings running at once. The reframing is the whole interview; two O(n log n) solutions fall out of it, and the tie-handling trips up the careless.

This problem rewards knowing that an in-order traversal of a BST yields sorted order. The signal is stopping early at the kth element rather than traversing the whole tree.

A row- and column-sorted matrix rewards one observation: start from a corner and kill an entire row or column each step. That turns an O(m*n) scan into O(m+n), and the choice of corner is not arbitrary.

Dropout is five lines, yet most candidates botch two of them: disabling it at inference and scaling the survivors so the expected activation stays constant. Inverted dropout is the convention they want to see.

Finding the longest run of consecutive integers without sorting separates candidates who reach for O(n log n) from those who see the hash-set trick. The signal is one loop invariant that kills redundant work. Here is the answer.

Validating a BST catches people who only check parent-child pairs and miss violations buried deep in a subtree. The signal is propagating min/max bounds or using in-order monotonicity. Here is the answer.

Course Schedule reduces to one question: is the prerequisite graph a DAG? The signal is spotting the graph framing and using topological sort or DFS to detect a cycle. Here is the answer.

RoPE is why modern LLMs extrapolate to longer contexts, and it's a rotation, not an addition. Coding it shows you actually understand how position enters attention. Here is the implementation.

A wrong split quietly inflates every metric you'll ever report, and the bugs are subtle: leaked groups, shuffled time, preprocessing fit on everything. Here is the split done right.

Many state-of-the-art training runs keep a shadow copy of the weights that's a moving average of the trajectory, and serve that instead of the final step. It's a few lines and a real accuracy boost. Here it is.

Almost every modern training run uses warmup-then-cosine, and getting it wrong destabilizes early training or wastes the tail. It's a closed-form function of the step. Here is the implementation and the why.

The textbook variance formula loses catastrophic precision on streaming data and cannot update incrementally. Welford's one-pass update is the standard for online feature normalization. Here it is.

Negative sampling in word2vec and recommendation draws millions of weighted samples; doing each in O(log n) is a bottleneck. The alias method makes each draw O(1) after an O(n) setup. Here it is.

When you shard a vector index or route requests across model replicas, naive 'hash mod N' reshuffles everything when N changes. Consistent hashing moves only a small fraction. Here is the implementation.

A Bloom filter answers 'have I seen this?' using a few bits per item instead of storing the key, trading a rare false positive for a huge memory win. Here is the build, the sizing math, and where it pays off in dedup and serving.

Floyd's tortoise and hare detects a cycle in O(1) space, but the part that separates strong candidates is the second phase: a short distance argument that pinpoints exactly where the loop begins. Here is the answer and the math.

Two lists that share a tail must intersect at a single node, but matching their lengths without counting is the elegant move interviewers want. Here is the two-pointer trick that makes the lengths cancel out, plus the answer.

The two-pass length-then-delete solution works, but the interview wants a single pass: a gap of N between two pointers, plus a dummy head that makes deleting the real head fall out for free. Here is the clean answer.

The diameter is not the tree's height, and the path need not pass through the root. The clean solution computes height and best path in one DFS, updating a global max at every node. Here is the answer and the subtlety candidates miss.

Preorder names the root; inorder splits the rest into left and right subtrees. Doing it naively is O(n squared); the strong answer uses a value-to-index map and a moving preorder pointer for O(n). Here is the answer and why both orders are required.

Negative subtrees, a path that bends through any node, and a return value that differs from the answer you track: this problem packs three traps into one DFS. Here is the clean O(n) solution and the reasoning that survives follow-ups.

The brute force is O(n squared) nested scans. A monotonic stack does it in one O(n) pass by keeping only the candidates that can still be a future answer. Here is the pattern and why each element is pushed and popped exactly once.

The naive triple loop is O(n cubed). Sort first, then fix one element and converge two pointers for O(n squared). The real test is deduplication: returning unique triplets without a set. Here is the clean template.

Merging intervals is the gateway to the whole interval family. The trick is sorting by start so overlaps land next to each other, then collapsing them in one linear sweep. Here is the pattern and the sweep-line generalization behind it.

Generating then filtering is O(2^(2n)) and wasteful. Backtracking with two counters builds only valid strings by enforcing the balance rule at every step. Here is the template and why the pruning conditions are exactly right.

The words encode a partial order between characters. Turn each adjacent pair into a directed edge, then topologically sort. The traps are the prefix edge case and detecting contradictions. Here is the full solution with cycle handling.

Words are nodes, one-letter edits are edges, and the shortest ladder is a shortest path in an unweighted graph, so BFS. The make-or-break detail is generating neighbors in O(26 * L) without scanning the whole dictionary. Here is the pattern.

Checking every subarray is O(n squared). Prefix sums plus a remainder-frequency map does it in one O(n) pass: two prefixes with the same remainder mod k bracket a divisible subarray. The trap is negative remainders. Here is the clean answer.

Evaluating arithmetic with precedence and nested parentheses is a parsing problem, and a stack handles both cleanly. The traps are operator precedence, multi-digit numbers, and the sign of truncated division. Here is a single-pass solution.

The rotting-oranges family of build screens: a state spreads outward one step per tick across a grid, and the spec gains rules each round. The signal is multi-source BFS by layers, not per-cell loops. Here is the design that absorbs each new test case.

A Fenwick tree answers prefix-sum queries and point updates in O(log n) with a single flat array and one bit trick. The signal is the lowbit operation and why it beats a plain prefix array under updates. Here is the answer.

A segment tree answers any associative range query (sum, min, max, gcd) with point or range updates in O(log n). The signal is the recursive split into covered, disjoint, and partial nodes, plus lazy propagation for range updates. Here is the answer.

Floyd-Warshall computes shortest paths between every pair of nodes in O(V cubed) with a tight three-line triple loop. The signal is the intermediate-node DP order and why k must be the outer loop. Here is the answer.

KMP matches a pattern in linear time by precomputing a failure function that skips redundant comparisons instead of backtracking the text. The signal is what the prefix function actually stores and why the text pointer never moves backward. Here is the answer.

Rabin-Karp slides a fixed-width hash over the text so each window costs O(1) to update, making it ideal for matching many patterns at once. The signal is the polynomial rolling hash, modular arithmetic, and why you still verify on a hash hit. Here is the answer.

Matrix exponentiation evaluates linear recurrences like Fibonacci at index n in O(log n) by raising a transition matrix to the nth power via binary exponentiation. The signal is building the transition matrix and squaring it. Here is the answer.

A from-scratch classic that tests recursion plus the split criterion math. The signal is computing impurity correctly, choosing the best threshold by information gain, and knowing the stopping rules. Here is a clean recursive implementation.

A from-scratch favorite that tests linear algebra fluency. The signal is centering first, using SVD instead of forming the covariance matrix, and reading variance off the singular values. Here is the implementation and the details interviewers push on.

A from-scratch test of the EM algorithm and soft clustering. The signal is the two alternating steps (responsibilities then weighted re-estimation), log-sum-exp for stability, and knowing how GMM generalizes k-means. Here is the implementation.

A from-scratch test of recurrent forward passes and backprop through time. The signal is the shared-weight recurrence, accumulating gradients across timesteps, and explaining the vanishing-gradient problem. Here is the implementation.

A from-scratch test of gated recurrence. The signal is wiring the forget/input/output gates and candidate correctly, separating cell state from hidden state, and explaining why the additive cell path fixes vanishing gradients. Here is the implementation.

A from-scratch test of modern optimizers. The signal is the moment estimates, bias correction, and the one change that matters: AdamW decouples weight decay from the gradient instead of folding it into the loss. Here is the implementation and why it wins.

A from-scratch test of how learning-rate schedules are actually computed. The signal is correct closed-form formulas, knowing when each schedule shines, and why decaying the rate helps convergence. Here is the implementation of the common ones.

A from-scratch test of how PyTorch actually works under the hood. The signal is building a computation graph during the forward pass, local derivatives per op, and a topological-order backward pass that accumulates gradients. Here is a minimal engine.

A from-scratch test of subword tokenization. The signal is greedy longest-match encoding against a fixed vocabulary, the continuation-prefix convention, and how WordPiece differs from BPE. Here is the implementation.

A from-scratch test of distributed input pipelines. The signal is partitioning data across workers with no overlap and no gaps, epoch-consistent shuffling with a shared seed, and handling the uneven-last-batch problem. Here is the implementation.

A sudden 22-point AUC jump is a gift and a warning. Junior candidates celebrate; strong ones get suspicious and know exactly which leakage checks to run before anything ships.

The most common ML fundamentals question, and a quiet seniority test: anyone recites the definition, but can you decompose the error and turn it into a concrete debugging plan?

The interviewer is testing whether you pick metrics for the problem or recite definitions. The real signal is knowing when accuracy and even AUC lie, and tying each metric to a decision.

Shipping a model is an experiment, and this question separates people who run A/B tests from people who p-hack them. The signal is pre-registering the metric, sizing the test, and resisting the urge to peek.

Imbalance shows up in fraud, churn, and abuse, and the naive answer (oversample, done) leaks data and inflates offline metrics. The signal is preferring cost-sensitive learning, fixing the metric, and resampling correctly. Here is the full toolkit with tradeoffs.

A staple stats question that doubles as a numeracy check. The signal is connecting MLE/MAP to regularization and nailing the base-rate calculation that surprises most people. Here is the intuition plus the worked numbers.

Optimizer questions test whether you understand what each one adapts and the subtle AdamW fix that the whole field now uses. The signal is the per-parameter adaptivity story plus why coupling weight decay to Adam was a bug. Here is that answer.

Tabular ML is won on features, and this question checks whether you handle the practical traps: target leakage in encodings, missingness that is informative, and fitting transforms on the wrong data. The signal is leakage-safe preprocessing. Here is the toolkit.

Recsys underpins half of applied ML, and this question checks whether you match approach to data availability and have a real answer for the cold-start problem that breaks naive systems. Here is the foundations answer.

A point-estimate win on a metric is not a real win. The signal is testing significance with the right paired method on a clean held-out set, weighing effect size, and confirming online. Here is how to compare models without fooling yourself.

A near-universal ML fundamentals question. Anyone can say 'L1 is lasso, L2 is ridge'; the signal is the gradient-and-geometry reason L1 drives weights to exactly zero and when you'd pick each. Here is that answer.

Ensembles dominate tabular ML, and this question checks whether you know they attack different parts of the error. The signal is bagging-reduces-variance vs boosting-reduces-bias and why GBMs win on tabular data. Here is the answer.

Trees are the atom of the ensembles that dominate tabular ML, so interviewers check you can explain how a split is chosen and how a single tree overfits. The Gini-vs-entropy part is a trap: candidates over-weight a choice that barely matters. Here is the answer.

Backprop is the algorithm that makes deep learning trainable, and the interviewer wants the chain-rule mechanics, not just 'it computes gradients.' The signal is the forward-then-backward flow and why it is efficient. Here is the answer.

BatchNorm is one of the most-asked deep-learning questions, and the trap is the train/inference difference. The signal is what it normalizes, why it stabilizes and speeds training, and why it switches to running statistics at inference. Here is that answer.

This question ties together why deep nets were hard to train and the cluster of tricks that fixed it. The signal is the multiplicative-gradient cause and naming the real fixes: ReLU, He/Xavier init, residuals, normalization. Here is the answer.

CNNs are foundational even as transformers rise, and this checks whether you understand why convolution suits images. The signal is parameter sharing and local connectivity, what pooling buys, and how the receptive field grows. Here is the answer.

Patches as tokens, global attention from layer one, and a weaker inductive bias than a CNN. The signal is naming the data regime where each architecture wins and why. Here is the answer interviewers score highest.

Forward noising is fixed, reverse denoising is learned, and the training loss is a plain noise-prediction regression. The signal is why that beats a GAN's minimax and what the VAE and U-Net each do in latent space. Here is the answer.

Senior DS loops test causal reasoning past the randomized A/B test. The signal is naming the quasi-experimental method that fits and stating the single assumption it lives or dies on. Here is the answer for when randomization is off the table.

A model can rank perfectly and still emit meaningless probabilities. The signal is knowing exactly when calibration matters, how to measure it, and the post-hoc fixes. Here is the answer most candidates miss.

Core statistics screening that catches people who run tests without understanding why they work. The signal is what the CLT actually licenses, and a correlation-vs-causation answer that uses Simpson's paradox to expose a confounder.

Self-supervision is how modern models pretrain on unlabeled data, the engine behind LLMs and modern vision. The signal is the pretext-task idea and the real difference between contrastive and masked-prediction objectives.

Multimodal is now table stakes, and this checks whether you understand how images and text reach a shared model. The signal is CLIP's contrastive alignment and how modern VLMs feed image features into an LLM's token space.

Sequence models are foundational and still asked, especially the gating that fixed RNNs and why attention won. The signal is connecting the vanishing-gradient story to the parallelism argument that let transformers ride the scaling wave.

GANs defined generative modeling for years, and interviewers want the minimax game, the failure modes (mode collapse, instability), and the precise reason diffusion displaced them for images while GANs still win on speed.

Autoencoders and VAEs underpin representation learning and generative modeling, including the VAE inside latent diffusion. The signal is the gap between a plain autoencoder and a variational one, and why you cannot backprop through sampling without the reparameterization trick.

Transfer learning is why you rarely train from scratch, and the real question is how much of the pretrained model to reuse versus adapt. The signal is a crisp decision grid over data size and task similarity, plus knowing when a low LR saves you from catastrophic forgetting.

Two simple classifiers that still appear in interviews because they test whether you understand assumptions and tradeoffs, not just sklearn calls. The signal is k-NN's laziness and curse of dimensionality versus Naive Bayes' independence assumption. Here is the answer.

Distribution questions test whether you can match a data-generating process to the right model. The signal is knowing what each distribution describes and a concrete use case, not memorizing PDFs. Here is the practical map.

Hypothesis testing underlies every experiment, and the signal is interpreting the p-value correctly (it trips most people) and picking the right test. Here is the rigorous, plain-language answer.

Labels are the expensive bottleneck in ML, and these two techniques attack it from different angles. The signal is knowing active learning chooses what to label while semi-supervised uses unlabeled data directly. Here is the answer.

Time series breaks the usual ML assumptions: data is ordered and correlated, so random cross-validation quietly leaks the future and inflates your score. The signal is decomposition, point-in-time features, and time-aware validation.

Ensembles win Kaggle and quietly power most production tabular models. The signal is explaining why combining models beats one (error decorrelation), and knowing exactly how the four techniques differ in what they fix.

Bandits optimize while they learn, shifting traffic toward winners mid-experiment instead of fixing it like an A/B test. The signal is the explore-exploit tradeoff, the three core algorithms, and knowing exactly when a bandit beats a clean A/B test.

Outlier detection shows up in data cleaning, fraud, and monitoring, and there is no single right method. The signal is matching statistical, distance, or model-based families to the dimensionality, distribution, and labels you actually have.

One model, several objectives, one shared backbone. The signal is knowing exactly why sharing helps (regularization, data efficiency) and the failure mode that sinks naive setups. Here is the answer.

The learning rate is the one hyperparameter that can NaN your run on step 50 or stall it forever. Modern training schedules it instead of fixing it. The signal is knowing exactly what warmup and decay each buy you. Here is the answer.

Ranking is not classification, so accuracy is the wrong tool. The signal is matching the rank-aware metric to the task: one right answer versus many, binary versus graded relevance. Here is the answer.

A staple that checks whether you know why ReLU dethroned sigmoid and how to match the output activation to the loss. The signal is the vanishing-gradient story told correctly. Here is the answer.

Everyone says 'gradient descent,' but the batch-size choice and momentum are the real interview content. The signal is the noise-vs-cost tradeoff across batch/SGD/mini-batch and why momentum earns its keep.

The loss defines what the model optimizes, and picking the wrong one quietly dooms it. The signal is matching the loss to the task and data (outliers, imbalance), not defaulting to MSE or cross-entropy on reflex.

Regression needs different metrics than classification, and each tells a different story. The signal is knowing what RMSE vs MAE vs R² vs MAPE actually measure and their failure modes around outliers, scale, and zeros.

Cross-validation gives a reliable performance estimate, but the wrong scheme leaks data and lies. The signal is matching the CV scheme to the data (stratified, grouped, time-series) and fitting preprocessing inside the fold.

These three get conflated, but they answer different questions: one is a preprocessing tool, two are eyeballing tools. The candidates who pass know exactly which structure each preserves and which it quietly throws away.

k-means is the reflex answer, but it quietly assumes round, equal-size clusters and demands you know k upfront. The signal is positioning each alternative by the exact assumption it removes, and knowing when to reach for it.

LDA is the supervised cousin of PCA, and the contrast is a favorite. The signal is knowing that LDA spends the labels to maximize class separation while PCA only chases variance, plus the C minus 1 dimension cap that catches most candidates off guard.

These four quantities underlie cross-entropy loss, decision-tree splits, distillation, and the KL penalty in RLHF. The signal is deriving them from one another and pointing to exactly where each shows up in a real training loop.

Preprocessing quietly decides whether a model trains well, and the right choice is model-dependent. The signal is knowing when scaling matters versus when it is wasted effort, and how to encode a feature with thousands of categories without blowing up your matrix.

Bad initialization makes deep nets fail to train at all, with activations that vanish or explode through depth. The signal is knowing why zero init is fatal and how to match Xavier or He to your activation function.

High-dimensional data breaks the intuitions and methods that work in two or three dimensions. The signal is naming the concrete effects (distances concentrate, data goes sparse, overfitting climbs) and the mitigations that actually move the needle.

Feature selection cuts noise, overfitting, and serving cost, but the three method families pull in different directions. The signal interviewers want is filter vs wrapper vs embedded and the conditions under which you reach for each.

XGBoost and LightGBM win most tabular problems, and interviewers want more than 'it's boosting.' The signal is the fit-to-residuals mechanism plus the engineering (regularization, histograms, second-order) that makes it both fast and accurate.

Pick the wrong distance metric and you quietly break kNN, clustering, and retrieval. The signal is knowing what each metric actually measures and matching it to the data: magnitude vs direction, sets, correlated features.

Contrastive learning is how modern embeddings (CLIP, sentence encoders, SimCLR) are actually trained. The signal is the pull-positives-push-negatives objective, the InfoNCE loss, and why the number and hardness of negatives makes or breaks quality.

Dropout is the classic neural-net regularizer, but the signal is whether you can explain why zeroing activations forces redundancy, and the train-vs-inference scaling bug that trips most candidates. Here is the answer.

Hand-labeling at scale is the bottleneck. Weak supervision generates labels programmatically instead, and the signal is whether you can explain how a label model denoises conflicting sources into probabilistic labels. Here is the answer.

Grid and random search ignore past results. Bayesian optimization learns from them, and the signal is whether you can explain the surrogate plus acquisition loop and name the exact condition where the sample-efficiency is worth it. Here is the answer.

Running an A/B test is easy; running a valid one is hard. The signal is naming the traps that produce confidently wrong conclusions and the fix for each. Here is the checklist a senior experimenter carries.

GNNs power recommendations, fraud, and molecule modeling by learning over graph structure. The signal is the message-passing mechanism, why k-hop matters, and why you keep them shallow. Here is the answer.

RL underpins RLHF, robotics, and recommendation, and interviewers want the core framing. The signal is the agent-environment-reward loop and the three things that make it harder than supervised learning. Here is the answer.

Data augmentation cheaply expands training data and regularizes models, but valid transforms differ by modality. The signal is the label-preserving constraint and why text is the hard one. Here is the answer.

Two cheap regularizers that fix overconfident classifiers. The signal is knowing that one softens the target and the other softens the input, and exactly why softer signals calibrate the model. Here is the answer.

Ranking is not regression: getting the absolute score wrong is fine, getting the order wrong is not. The signal is the pointwise/pairwise/listwise split and why pairwise (LambdaMART) is still the practical default. Here is the answer.

Classic bias-variance says bigger models eventually overfit, yet deep nets keep improving past the point where they memorize the data. The signal is explaining the second descent and why over-parameterized models generalize. Here is the answer.

Many real problems give you confirmed positives but never confirmed negatives, only unlabeled data. The shortcut everyone reaches for quietly biases the model. The signal is naming the regime and its fix. Here is the answer.

Some models are natively multiclass; others need a wrapper to stretch a binary classifier to K classes. The signal is knowing the three strategies and exactly when each one wins. Here is the answer.

A model can rank perfectly and still lie about its probabilities; its 0.9 may be right 70% of the time. The signal is knowing the three fixes, which data size each needs, and how to measure the gap. Here is the answer.

The textbook reflex ('drop one') is usually the wrong instinct, and whether collinearity matters at all hinges on your model and what you need from it. This is the answer that separates rote from understanding.

Everyone can recite 'linearity and normality.' Few can say which assumption matters for predictions versus inference, how to spot a violation in a residual plot, and what to actually do about it. Here is that answer.

A pre-existing gap between your groups means randomization or instrumentation is broken, and the whole experiment is suspect. Strong candidates name sample-ratio mismatch on sight. Here is the full diagnosis and the disciplined response.

A 92% aggregate accuracy can hide 60% on the segment that matters most. Average metrics are exactly where these failures hide. Here is how to surface them and the menu of fixes that actually map to the cause.

'The model won't train' has a short list of usual suspects, each with a distinctive loss-curve signature. The shape of the curve names the bug before you touch a single hyperparameter. Here is how to read it.

If labels only exist for the cases you already acted on, the model learns a distorted world: great offline, blind to everyone you never saw. Worse, its own decisions pick the next labels. Here is how to spot it and counter it.

Defaulting to 0.5 is leaving money or safety on the table. The right cutoff comes from the cost of each error and the operating constraint, not the model. Here is how to set it deliberately.

The built-in importance scores from XGBoost can rank a random ID above a vital feature, and stakeholders will make decisions on that bar chart. Here is why default importance lies and what a careful answer reports instead.

Accuracy ties are common, and accuracy is rarely the deciding factor anyway. The real tiebreakers are the ones a junior candidate forgets to mention. Here is the checklist a staff engineer runs before shipping either one.

'The model got worse' has three very different causes and three different fixes. Retraining a model that's actually broken by a pipeline bug just bakes in garbage. Here is the triage order.

Most real datasets have wrong labels, and they cap your accuracy invisibly. Cleaning all of it by hand doesn't scale. Here is how to find the bad labels and train robustly around them.

The idea that models, like students, learn better from easy examples first is intuitive, and sometimes it works and sometimes it does nothing. Knowing when it pays off is the real signal. Here is the honest answer.

Metric learning lives or dies on the pairs you feed it. Random negatives teach almost nothing, and the margin and mining strategy decide whether the embeddings are any good. Here is how the choices interact.

Boxes and labels, not one tag per image. The signal is choosing two-stage vs one-stage by your latency and accuracy budget, knowing what NMS and anchors actually do, and reporting mAP correctly. Here is the answer interviewers score highest.

Cross-entropy is the start, not the answer. The signal is matching the loss to the task structure: focal for detection's background flood, IoU-family for box overlap, Dice for imbalanced masks. Here is how to reason about it.

Semantic labels every pixel by class; instance separates each object. The signal is knowing that one cannot count overlapping objects, why Mask R-CNN adds a mask branch on RoIAlign, and what RoIAlign fixed. Here is the answer.

A million rows does not automatically mean deep learning. The signal is reasoning from dimensionality, data type, inference latency, and label budget, then noting where KNN quietly survives as approximate nearest neighbor. Here is the answer.

Resampling, reweighting, and focal loss attack class imbalance from different angles, and each has a real downside. The signal is matching the method to the model and metric, not blindly oversampling. Here is the breakdown.

IV is the causal tool people name but can't defend. The signal is articulating the two conditions an instrument must satisfy, why one is testable and one is not, and what 2SLS actually estimates. Here is the answer.

Propensity scores promise to mimic an experiment from observational data. The signal is knowing what they can and cannot fix, the overlap trap, and why IPW blows up. Here is the answer that separates careful candidates.

CUPED can cut the sample size an A/B test needs by half without touching validity. The signal is explaining why subtracting a pre-experiment covariate reduces variance but never biases the estimate. Here is the answer.

Standard A/B math assumes one user's treatment doesn't affect another's outcome. In marketplaces and social networks that assumption breaks and your estimate is biased. The signal is naming SUTVA and the right randomization unit. Here is the answer.

Prophet treats forecasting as curve-fitting trend plus seasonality plus holidays, not as a stochastic process. The signal is knowing why that design wins on business data and where it quietly fails. Here is the answer.

There is no single fairness number, and a famous result proves you cannot satisfy the main three together. The signal is defining each metric precisely and explaining the impossibility, not picking one blindly. Here is the answer.

Sessionization separates people who reach for a self-join from people who know LAG plus a running sum. Here is the two-pass pattern that scales to billions of events, plus the edge cases interviewers push on.

Pipelines fail and retry; the real question is whether a retry corrupts your data. The signal is idempotent writes (not 'exactly-once delivery') plus watermarks for late data. Here is how to make re-runs safe without double-counting.

Top-N-per-group is the window-function question every data round asks, and the trap is RANK vs ROW_NUMBER vs DENSE_RANK. The signal is choosing the right ranking function for ties and knowing window frames. Here is the pattern and the tie nuance.

A hard streaming-systems question: dedup at high throughput with bounded state. The signal is a tiered state design (probabilistic filter in front of durable state) plus watermark-driven eviction. Here is the architecture that does not OOM.

A pipeline-design question that rewards matching the architecture to the freshness requirement, not chasing real-time for its own sake. The signal is the latency-vs-complexity tradeoff and knowing why Kappa emerged to kill Lambda's dual codebase. Here is the decision.

A data-warehouse fundamentals question that separates people who model for analytics from people who only know OLTP normalization. The signal is facts vs dimensions, the star schema, and why analytics denormalizes where transactional systems normalize.

The difference between a query that scans a terabyte and one that scans a gigabyte is usually layout, not the engine. The signal is partition pruning, columnar formats, and the small-files problem.

Syncing a warehouse with a live OLTP database is a CDC problem, and the naive 'full reload nightly' or 'query by updated_at' answers have real holes. The signal is log-based CDC and idempotent merges.

Bad data silently poisons everything downstream (dashboards, models, decisions), and 'it ran without error' is not 'it's correct.' The signal is the categories of checks and failing loud at the boundary.

Query tuning is a core data-engineering skill, and the signal is reading the execution plan before changing anything, then applying the right fix. Here is the diagnostic method that separates guessers from engineers.

Dedup is a daily data-engineering task and a common screen. The signal is ROW_NUMBER to keep the right record (not just DISTINCT), defining what 'duplicate' actually means, and handling near-duplicates at scale.

Pivoting (rows to columns) is constant in analytics and reporting, and the portable trick is conditional aggregation, not a vendor PIVOT clause. The signal is the CASE-inside-aggregate pattern and knowing its one real limit.

The canonical product-analytics SQL question. The signal is grouping users by signup cohort, computing each activity's period offset, and counting distinct returners per offset to build the retention triangle.

Joins are the heart of SQL, and the real signal is precise semantics: semi/anti joins (EXISTS / NOT EXISTS), the NOT IN null trap, and the fan-out bug that silently doubles your sums.

The subtlety in rolling aggregates is the window frame clause, especially ROWS vs RANGE. The signal is the correct frame for a moving window and the tied-rows gotcha that corrupts running totals.

Means lie on skewed data like latency and spend, and SQL has dedicated functions for the truth. The signal is PERCENTILE_CONT/DISC for exact quantiles, NTILE for bucketing, and why p50/p99 beat the average.

A new column is safe; a rename silently corrupts every dashboard downstream. The signal is knowing which changes are compatible, enforcing data contracts in CI, and running expand-contract migrations for the breaking ones.

You cannot wait for all the data, and buffering an unbounded stream will OOM. The signal is windowed joins with watermarks for stream-stream, lookup joins for stream-table, and a clear story for late and out-of-order events.

Set operations and NULL semantics trip up even senior engineers. The signal is UNION vs UNION ALL and its dedup cost, plus three-valued logic where NULL equals nothing, not even itself.

The top cause of mysteriously slow Spark jobs: one partition does most of the work while the rest idle. The signal is reading the symptom (a few straggler tasks) and reaching for the right fix, salting, broadcast, or AQE.

Time-based analysis is everywhere in analytics, and time zones and bucketing are the common traps. The signal is DATE_TRUNC for bucketing, interval math, and storing UTC.

Correlated subqueries read naturally but can execute per row (O(n squared)), a classic slow-query cause. The signal is recognizing them and rewriting as joins or window functions.

Reporting often needs subtotals and grand totals across several dimensions, and stitching that with many UNIONs is wasteful. The signal is GROUPING SETS, ROLLUP, and CUBE in one pass.

Exact COUNT(DISTINCT) over billions of rows tracks every unique value, so it gets slow and expensive. The signal is knowing when a sketch like HyperLogLog buys huge memory and speed wins for a sub-percent error budget.

The single most common way SQL leaks the future into a training set is a careless join to a feature table. Point-in-time correctness is the fix, and it is an as-of join. Here is how to write it.

Funnels look like simple counts until you require the steps happen in order and within a window. Counting each step independently overstates conversion. Here is the ordered-funnel query.

Consecutive-run problems (active streaks, uptime windows, price-stable periods) all reduce to one trick. Once you see the row-number difference, they collapse to a window function and a group-by.

Building a training set from a billion-row table means sampling, and the naive 'ORDER BY random() LIMIT n' sorts the whole table. Stratified sampling needs more care still. Here are both, done efficiently.

Bucketing turns a skewed continuous feature into something a model (or a report) handles well, and the choice between equal-width and equal-count bins changes everything on skewed data. Here is both in SQL.

Production drift monitoring often lives in the warehouse, not a fancy tool. Comparing two distributions in SQL with a metric like PSI is a few CTEs. Here is how to compute it.

An org-chart recursion walks a tree, but a general graph has multiple paths and back-edges. The signal is accumulating the visited path to prune cycles and ranking paths by cost to get the shortest one.

Most candidates only know 'use Debezium.' The signal is comparing three CDC mechanisms by how they capture deletes, load the source, and handle ordering, plus the snapshot-to-stream stitching that trips up real deployments.

Anyone can say 'filter and dedup.' The signal is the funnel in cost order, the MinHash/LSH banding math, and knowing the shuffle at billions of documents is what actually costs you, plus benchmark decontamination people forget until their eval numbers get challenged.

This is two questions hiding as one: how would you have caught it before full rollout, and how do you debug it now. Answer both and you signal senior judgment. Here is the staged-rollout and root-cause playbook.

The modal ML system design round. The structure interviewers reward is the funnel: candidate generation then ranking then re-ranking, with the right model at each stage and an honest plan for cold start, freshness, and feedback loops. Here is that structure.

Extreme class imbalance plus a hard latency budget plus an adversary who adapts. The signal is handling imbalance honestly, choosing the operating point from costs, and designing for the feedback loop. Here is the end-to-end design.

Models fail silently; the question is whether you would know. The signal is monitoring the right layers (operational, data, prediction, outcome) and alerting on drift without drowning in false pages. Here is the system and the metrics that matter.

Multimodal search tests whether you understand shared embedding spaces and the retrieve-then-rank pattern under a real catalog's scale and freshness. The signal is CLIP-style joint embeddings plus hybrid retrieval and honest relevance evaluation. Here is the design.

Moderation is a multi-stage classification problem with brutal tradeoffs: false negatives cause real harm, false positives censor legitimate users, and the adversary adapts. The signal is the tiered pipeline, per-severity precision/recall calibration, and human-in-the-loop. Here is the design.

Ads ranking is where calibrated probabilities meet hard latency and money. The signal is knowing that CTR must be calibrated (not just ranked), the feature and serving design, and the auction context. Here is the design that goes beyond 'train a classifier.'

Resume screening is high-scale and high-stakes: a legally sensitive, bias-prone decision about people. The signal is balancing throughput with fairness, human oversight, and explainability, not an LLM ranking resumes. Here is the design.

Dynamic pricing blends demand forecasting, optimization, and real-time serving, with fairness and trust constraints. The signal is the predict-then-optimize structure plus guardrails against perverse outcomes. Here is the design.

E-commerce search is where retrieval, ranking, and business goals collide, and it carries a hard exact-match requirement (brand, size, SKU). The signal is query understanding plus hybrid retrieval plus business-aware ranking. Here is the design.

A GPU-heavy generative-audio system with text/genre/lyric conditioning, minutes-long coherence, and copyright constraints. The signal is the generation pipeline plus the async, cost, and safety layer that makes it shippable.

Video generation is image generation plus the brutal constraint of temporal consistency, and it is the most GPU-intensive generative task there is. The signal is diffusion over spacetime plus a serving design that survives multi-minute jobs.

Notification systems fail by spamming users into muting them, and a mute is a permanently lost channel. The signal is treating it as a per-user send/hold/batch/suppress decision optimized for long-term trust, not for clicks or volume.

How large recommenders and search pull candidates from millions of items in milliseconds. The signal is why the user and item towers are separate, how that unlocks precomputed embeddings plus an ANN index, and where ranking takes over.

Every keystroke needs ranked suggestions back in tens of milliseconds, at search-engine QPS. The interesting part is the data structure and what you precompute, not the query path. Here is the design that survives the latency budget.

PYMK is graph recommendation at billion-node scale. The interviewer is screening for one instinct: do you generate candidates from the social graph, or naively try to score all pairs? Here is the design that survives the follow-ups.

Feed ranking is multi-objective ML at massive scale with a brutal serving constraint. The interviewer wants to see whether you optimize a value model toward long-term satisfaction, or fall into the CTR trap that breeds clickbait. Here is the design.

Predict, value, and bid on a single impression in about 10ms, billions of times a day. The signal is the pCTR-to-bid pipeline under a brutal latency budget, plus why calibration and budget pacing decide whether you make or lose money.

Image search is embedding retrieval at billion-image scale. The tell of a strong answer is knowing when a vision-only encoder suffices and when you need a CLIP-style shared space, plus how the ANN index actually serves the query. Here is the design.

Spam detection is adversarial, imbalanced classification under a low-latency bar. The interviewer is watching for whether you set the threshold by cost asymmetry and lean on signals spammers cannot fake. Here is the design.

Query understanding is the front end of search that turns a raw query into structured intent. Get it wrong and even a perfect index returns garbage. The signal is the pipeline and how you measure it. Here is the design.

Online models need features that are fresh and computed exactly the way training computed them. The classic failure is training-serving skew that silently degrades the model. The signal is the online/offline split with one shared definition. Here is the design.

New users and new items have zero interaction history, so collaborative filtering has nothing to work with. The candidates who pass treat it as a lifecycle, not a single trick. Here is how.

Translation is seq2seq generation served across hundreds of language pairs under tight latency and quality budgets. The signal is dodging the quadratic pair explosion, rescuing low-resource languages, and serving it cheaply. Here is the design.

Recognizing walking, driving, or a fall from a sensor stream is windowed time-series classification, usually on a battery-bound device. The signal is windowing, temporal smoothing, and the on-device constraints most candidates skip. Here is the design.

Matrix factorization is the classic collaborative-filtering method and the conceptual root of modern embedding-based recsys. The signal is the latent-factor idea and how it leads to two-tower/neural models.

Recommending events (or jobs) breaks normal recsys in two places: items expire and items are local. The signal is how you handle perishability, geography, and the fact that every item is a cold-start item. Here is the design.

Tagging articles at scale is bread-and-butter ML, but the easy version fails three ways: it is multi-label, the taxonomy is hierarchical and keeps growing, and rare classes hide behind aggregate accuracy. The signal is the model choice plus how you handle all three.

A recommender trains on data its own past recommendations produced, so it learns to confirm its own beliefs. The signal is recognizing the loop, naming the biases it breeds, and knowing the exploration and debiasing fixes that break it.

Labeled data is the fuel for ML, and a labeling platform lives or dies on quality control. The signal is the workflow plus the quality math: consensus, gold honeypots, inter-annotator agreement, and active learning to spend the budget where it counts.

Detecting and boxing objects at scale comes down to one driving tradeoff plus a handful of CV specifics most candidates fumble: the detector family, NMS, focal loss, and why mAP not accuracy is the metric. Here is the design.

Classic distributed-systems tradeoffs show up in ML infra with an ML twist: stale features and eventually-consistent reads have model-accuracy consequences, not just correctness ones. Here is how to reason about it.

Anthropic's signature opener. The naive answer (one server pushes to N clients) is bottlenecked by the source uplink and scales linearly. The answer interviewers reward turns receivers into senders so the swarm's capacity grows with its size. Here is that reasoning.

The Dynamo question. Interviewers want three decisions made cleanly: how you partition keys so adding nodes does not reshuffle everything, how you replicate for durability, and where you land on the CAP spectrum. Here is how to reason about all three with concrete quorum math.

Looks simple until the follow-ups: which layer do you balance at, how do you avoid sending traffic to a dead server, and how is the load balancer itself not a single point of failure? Interviewers want the algorithm, the health-check loop, and the high-availability story.

A deceptively rich warm-up. The interesting decisions are how you generate short, collision-free, unguessable keys at scale, and how a read-heavy 100:1 workload shapes the storage and caching. Here is the clean design with capacity math.

The observability infrastructure question. Interviewers want you to separate the three telemetry types, choose storage that survives massive write volume, and design alerting that fires fast without drowning on-call in noise. Here is the pipeline with retention and cardinality math.

The distributed ordering question, the conceptual core of Kafka, replicated logs, and consensus systems. Interviewers want to see you confront the hard truth that physical clocks lie, then reach for the actual tools: a single sequencer, consensus, or logical clocks. Here is that progression.

The real-time messaging classic. Interviewers want the delivery path (persistent connections, not polling), the fan-out strategy that does not collapse on large groups, and an ordering and delivery-guarantee story. Here is that design with the fan-out tradeoff that separates seniors.

This is the question to design the scoreboard, not the player. A strong answer separates offline gates (AUC, calibration, NDCG) from the online verdict (A/B with revenue, user, and advertiser guardrails), and adds counterfactual replay so you can trust a model before it ever serves a real auction.

A news feed is not a generic recommender. The clock is a first-class signal: a story that mattered this morning is noise by tonight. The interviewer wants recency decay, an engagement-versus-quality value model, and an answer for filter bubbles, not just retrieval plus ranking.

Truth is not a label you can train on cheaply, and adversaries adapt the moment you ship. A strong answer fuses content, graph, and behavioral signals, puts humans in the loop where precision matters, and treats adversarial drift as a permanent operating condition, not a one-time training problem.

Video search is image search with a time axis, and the time axis is the whole problem. A strong answer embeds frames, pools them into scene vectors, indexes with ANN at billion scale, and explains how a multimodal query (clip, text, or both) finds the right moment, not just the right video.

Moderating billions of items a day with an LLM on every one is too slow and too expensive. Learn the tiered funnel (cheap classifier then LLM then human), how to tune thresholds for precision versus recall, and how to keep up with adversaries.

Personalizing an LLM is a retrieval and memory problem, not a per-user fine-tune. Learn how to assemble user context at request time, manage long-term memory without bloating the prompt, and respect privacy and the right to be forgotten.

Live speech-to-speech translation is a latency race across three models (recognize, translate, synthesize) while audio keeps streaming in. Learn the streaming pipeline, how to commit partial results without flip-flopping, and the tradeoff between latency and translation quality.

A celebrity posts and 50M followers need a notification. The hard parts are not sending one message, they are fan-out strategy, deduplication, channel routing, and not flooding users. Here is the architecture that handles both the long tail and the viral spike.

A cache is easy until you scale it across many nodes. Then you face consistent hashing, eviction policy, the thundering herd on a cache miss, hot keys, and how stale you can tolerate. Here is the design that holds up under real traffic.

Crawling a few pages is trivial. Crawling the web means a URL frontier with priority, politeness per host, dedup at billions of URLs, and defenses against spider traps and infinite content. Here is the crawler architecture that scales without getting your IPs banned.

Showing the top 10 is easy. Showing a player their global rank among 50M others, updated live, is the real challenge. The answer hinges on the right data structure, sharding, and approximate ranks for the long tail. Here is how leaderboards scale.

When a page loads, an ad slot is auctioned to dozens of bidders and a winner is chosen before the page finishes rendering, all in under 100ms. The design covers the auction mechanism, the brutal latency budget, budget pacing, and fraud. Here is how an exchange runs millions of auctions per second.

A trustworthy experiment platform is far more than a 50/50 split. The signal is consistent assignment, exposure logging, statistical rigor, and guardrails that survive peeking and sample-ratio mismatch. Here is the design.

Models are only as good as their labels, and noisy annotation silently caps performance. The signal interviewers want is a measured quality process, not a louder collection effort. Here is the answer.

Same model, two answers: clean offline, ugly online. The culprit is almost always a feature computed differently in the two paths. Here is the diff-based hunt that localizes it to a single column.

A fraud model fed a feature an hour behind is half-blind, but recomputing everything in real time burns money you don't need to spend. Freshness is a per-feature decision on a real cost curve. Here is how to manage it.

If you wait for labels to measure accuracy, you find out a model broke a month after it broke. Production monitoring has to work before the truth arrives. Here is what you watch instead.

A data engineer renames a column and three production models silently start scoring on garbage. Features are a shared contract, and changing them needs the same care as a breaking API change. Here is the discipline.

An alert that fires constantly gets muted, and a model that fails with no alert is worse. Good ML alerting is a design problem with the same principles as SRE, plus ML-specific twists. Here is how to get it right.

Every team reinvents a spreadsheet of training runs and then drowns in it. The interviewer wants the platform that ingests runs, params, metrics, and artifacts at high write volume, links them by lineage, and makes thousands of experiments comparable, which is a different system from a model registry.

Three drifts get casually lumped together, but they have different detectability and different fixes. The interview signal is knowing which you can catch without labels and which one genuinely demands a new model.

You cannot wait for ground truth to tell you the model broke. Prediction-side monitoring catches failures in minutes, not weeks, if you know which output signals actually move first.

ML incidents are slipperier than service outages: nothing crashed, the model was just wrong. The strong answer covers detection, mitigation, and a blameless postmortem that fixes the system, not the person.

Production usage can become your best source of training data, or a self-reinforcing trap. The signal is knowing how to capture clean feedback and how to break the loops that quietly corrupt the model.

FSDP and ZeRO let you train models too big for one GPU without splitting the compute. The signal is the ZeRO stages (what gets sharded at each level) and the communication you pay for it. Here is the answer.

Distributed training is bottlenecked by GPU-to-GPU communication, and these collectives are how the data moves. The signal is what each one does and which parallelism strategy depends on it. Here is the answer.

Activations, not just weights, can dominate training memory, and gradient checkpointing is the standard fix. The signal is the precise trade: recompute activations in the backward pass instead of storing them. Here is the answer.

Batch size is a training knob with non-obvious effects on speed, memory, and generalization. The signal is the large-batch tradeoffs and the learning-rate-scaling and gradient-accumulation tricks that make scaling actually work.

OOM is the most common wall in deep learning, and 'buy a bigger GPU' is the weakest answer. The signal is naming the memory consumers, knowing which one dominates, and matching the right lever to it.

nvidia-smi showing 100% can hide that you are using a fraction of the hardware's real compute. The signal is MFU (useful FLOPs vs peak) and the gap between 'the GPU is busy' and 'the GPU is efficient'.

Beyond FP16/BF16, FP8 is the next rung on the precision ladder, and the latest tensor cores compute on it natively. The signal is knowing the two formats, what FP8 actually buys, and why dynamic range forces careful scaling.

Pruning removes redundant weights to shrink and sometimes speed up a model. The signal is the structured-versus-unstructured split and why only structured pruning reliably wins latency on commodity hardware.

Paying for accelerators that idle half the time is the most common waste in ML training, and the instinct to add more GPUs makes it strictly worse. The interviewer wants the profiling discipline that finds what is starving them.

Linear scaling is the marketing number; the real curve bends early for reasons that are physics, not bugs. Here is where the speedup goes and how to claw it back.

A 256-GPU job that wedges with no error at 3am is a special kind of pain. The causes are a short, recurring list. Here is the systematic way to find which one bit you.

At billion-parameter scale, training can be cruising and then the loss jumps and never comes back, burning a fortune in compute. The causes and the playbook are well-known to the few who've done it. Here it is.

No single parallelism dimension trains a 175B model on its own. The art is layering three of them onto the right hardware so the expensive communication stays on the fast links. Here is how to size each axis.

ZeRO and FSDP both shard training state across data-parallel ranks, but the magic is in when each shard is gathered and freed. Get the memory accounting and the all-gather timing and you can size any run.

Full activation checkpointing saves memory but burns a flat 30% extra compute. Selective recomputation recovers most of that by only recomputing the cheap, memory-heavy operations. Here is how to pick what to recompute.

The collective communication in distributed training is pure overhead unless it runs while the GPU computes. Hiding it is the difference between 30% and 55% MFU. Here is how the overlap actually works and how you confirm it on a trace.

MoE buys you more parameters for the same FLOPs, but the gating network, the load imbalance, and the all-to-all shuffle introduce failure modes dense models never see. Here is the training playbook.

RLHF with PPO is not one model training, it is four models in the loop at once, three of them on the GPU during every step. The memory math and the generation bottleneck are what trip people up.

On a thousand GPUs something fails every few hours. The question is not whether you lose a node, but how many GPU-hours you lose when you do. Checkpoint frequency, sharded writes, and fast restart decide that.

When all-reduce is the bottleneck, the cause is almost always a misconfigured path: traffic on the wrong link, a downed NIC, or a topology NCCL never discovered. Here is the systematic way to find the lost bandwidth.

FP8 promises to roughly double training throughput over BF16, but with only a handful of mantissa bits the numerics are unforgiving. Per-tensor scaling and a selective recipe are what make it converge.

Data, tensor, and pipeline parallelism all leave one sequence's activations on one device, so 200k+ token training hits a wall none of them can fix. The fourth axis shards the sequence itself, and the interview lives in the communication math.

Toxicity detection looks like vanilla text classification and is anything but: context flips labels, adversaries evolve weekly, and naive models flag dialects as hate. The signal is naming those failure modes and designing the human-in-the-loop system around them.

Privacy in ML is a toolbox, not a single switch, and each tool defends a different threat. The signal is mapping DP, federated learning, confidential computing, encryption, and minimization to what they actually protect, and knowing they compose.

Deleting a row is easy. Deleting a person's influence from embeddings, caches, derived datasets, and a trained model is not. GDPR and CCPA require it anyway. Here is the plan that survives an audit.

A backdoored model behaves perfectly until it sees a secret trigger, then flips. The signal is explaining why clean test accuracy never reveals it, and what detection actually buys you when the trigger is unknown.

Gibberish-looking token strings appended to a prompt can reliably break refusals, and they transfer between models. The signal is explaining the gradient search that finds them and why output checks beat input pattern matching.

Epsilon is the privacy knob, and shipping DP means choosing a number, defending it, and tracking what each query spends. The signal is treating epsilon as a finite budget with composition, not a magic constant. Here is the answer.

Federated learning keeps data on-device, but it costs you accuracy, debuggability, and engineering complexity. The signal is naming when those costs are justified and when a simpler centralized pipeline with DP wins. Here is the answer.

Consent and retention decide whether you can legally train on data and how long you keep it. The signal is purpose limitation, granular consent, enforceable TTLs, and a story for deleting data already baked into a model. Here is the answer.

This question is not about whether you failed; everyone has. It is a test of ownership, debugging rigor, and honesty under pressure. Here is the structure that turns a failure story into a hire signal, and the traps that turn it into a flag.

Every loop has this question and most answers fail in one of two ways: too combative or too much of a pushover. The signal is data-driven persuasion plus genuine disagree-and-commit. Here is the arc that lands and the traps that sink it.

Applied AI work is ambiguous by nature, and interviewers (Meta's 'embracing ambiguity,' Anthropic's deliberately under-specified problems) test whether you freeze or drive. The signal is structured progress under uncertainty. Here is the arc that lands.

Senior applied-AI work is mostly influence, not command: you need other teams to prioritize your dependency, adopt your approach, or change a plan you do not own. The signal is persuading with evidence and shared goals. Here is the arc that demonstrates it.

Applied AI rarely waits for certainty, and companies (Amazon's Bias for Action, NVIDIA's speed) test whether you can act decisively without freezing or being reckless. The signal is a reversible, well-reasoned call with a plan to validate. Here is the arc that lands.

Strong applied-AI engineers are known for NOT reaching for ML when they shouldn't. The signal is judgment: ML earns its complexity only under specific conditions, otherwise rules and heuristics win.

Applied AI engineers who think in business value, not just model metrics, stand out. The signal is tying the feature to a business metric, modeling total cost, and being willing to kill it. Here is the framework.

This question prizes intellectual honesty and a data-driven mindset over a polished win. The signal is running a real experiment, accepting a result you did not want, and acting on it. Here is the arc that lands at experiment-heavy cultures.

This question tests whether you can deliver under pressure without missing the date or shipping something broken. The signal is ruthless scoping and protecting quality where it matters, not all-nighter heroics. Here is the arc that lands.

Ownership is a top leadership signal: taking responsibility for an outcome no one assigned you. The signal is spotting a gap, owning it end to end including the unglamorous parts, and driving measurable impact. Here is the arc that lands.

Applied AI is persuasion backed by evidence, and this tests whether you can move a skeptic with data rather than rank or opinion. The signal is understanding their concern, bringing the exact evidence that addresses it, and speaking in their terms.

Simplification is a senior signal: anyone can add complexity, removing it safely takes judgment. The interviewer wants a real case where you cut complexity, separated essential from accidental, and it paid off measurably.

This probes whether you handle interpersonal friction with maturity, not who was right. The signal is seeking to understand the other side, anchoring on the shared goal, and preserving the relationship through the disagreement.

Customer obsession is Amazon's first Leadership Principle and a universal product signal. The interviewer wants evidence you start from the user's real need, not the technology, and advocate for the harder right solution over the easy one.

A depth probe disguised as a story. They want to watch you reason through genuine difficulty and keep going when pushed three levels down. The signal is a real hard problem, a systematic attack, and a root cause you own cold.

A soft-sounding question that quietly screens out low-effort candidates and tests real fit. The signal is specific, researched reasons tied to your trajectory, not flattery. Here is the structure that lands.

This probes whether you can move fast under uncertainty without being reckless. The signal is a calculated, reversible risk taken with judgment. Here is the arc that lands, especially at Amazon.

How you deliver bad news reveals maturity and trustworthiness. The signal is communicating early, honestly, and with a plan, not hiding or sugarcoating it. Here is the arc that lands.

This probes resilience: can you push through sustained difficulty without giving up or burning out? The signal is staying the course through setbacks with adaptation, not stubbornness. Here is the arc that lands.

This tests self-awareness and growth: can you critique your own work honestly and extract lessons? The signal is a genuine, specific improvement you'd make, owned without defensiveness. Here is the arc.

Applied AI work is cross-functional: you ship with PMs, engineers, and domain experts. This probes whether you collaborate effectively across boundaries. The signal is bridging different goals and languages toward a shared outcome. Here is the arc.

The classic question that screens for self-awareness and honesty, not for a perfect non-answer. The signal is a real weakness plus what you're actively doing about it. Here is how to answer it well.

Giving hard feedback well is a maturity and leadership signal. Interviewers want candor delivered with care that actually changed behavior, not avoidance and not bluntness. Here is the arc that lands.

Real work means more demands than time, often from stakeholders who each think theirs comes first. This probes how you prioritize and communicate. The signal is prioritizing by impact transparently, not just working harder.

AI moves fast, requirements shift, and tools change monthly. This probes adaptability. The signal is embracing the change, learning quickly, and adjusting course productively, not resisting it. Especially relevant for applied AI.

Saying no well is a senior signal: protecting quality, scope, or users without damaging the relationship. The signal is principled pushback with reasoning and an alternative, not just refusal. Here is the arc.

Defining the right metric is a high-judgment skill, and in AI the offline number and real impact routinely diverge. The signal is a metric tied to outcomes and guarded against gaming. Here is the arc.

The 2025-2026 AI-fluency screen. Interviewers want a concrete daily workflow, a real verification discipline, and a story where the AI was wrong. Vague enthusiasm fails; here is what passes.

A favorite practical screen at the labs, because it tests concurrency, time handling, and judgment in 20 lines. The trap is the background thread that wastes CPU. Here is the lazy-refill version interviewers want, plus the follow-ups.

Applied AI is half translation. Interviewers ask this to see whether you can drop the jargon, frame tradeoffs in business terms, and earn trust with people who control the roadmap. Here is how to show that skill in the answer itself.

Every chat model's 'temperature' and 'top_p' knobs are a few lines of logit surgery. Implementing them shows you understand decoding, not just call it. Here is the from-scratch version.

Dijkstra finds single-source shortest paths in O((V+E) log V) with a min-heap, the backbone of routing and network latency problems. The signal is why it needs non-negative weights and how lazy deletion keeps the heap simple. Here is the answer.

Bellman-Ford handles negative edge weights that break Dijkstra and detects negative cycles, the basis of currency-arbitrage problems. The signal is the V-1 relaxation rounds and the extra Vth round that flags a negative cycle. Here is the answer.

Kruskal builds a minimum spanning tree by sorting edges and adding the cheapest that does not form a cycle, using union-find for the cycle check. The signal is the greedy cut property and why disjoint-set is the right tool. Here is the answer.

This is a full interview round (OpenAI's project deep dive, Amazon's Tech Talk), not small talk. It is where they test whether you actually understand the work you claim. The signal is depth, your specific contribution, and the tradeoffs you can defend. Here is how to prepare and deliver it.

A staple ML-coding screen. Writing the loop is easy; the signal is k-means++ init, a real convergence criterion, and knowing the failure modes. Here is the tight implementation plus the follow-ups interviewers always ask.