A Machine Learning Engineer at Pinterest plays a pivotal role in shaping how more than 600 million users discover, save, and act on visual inspiration. With an extraordinary dataset comprising over 300 billion saved ideas ("Pins"), the engineering challenge is not just about applying off-the-shelf algorithms. It is about building highly specialized, large-scale recommendation systems, deep learning models, and real-time streaming pipelines that can process massive graphs of user interactions.
The impact of this role is directly visible across the core product surfaces. Whether it is personalizing the Homefeed, optimizing search retrieval, enhancing shopping recommendations, or maximizing revenue through ads monetization, machine learning is the engine that drives the business. As an engineer on this team, you will work on the monetization stack, user modeling, or content understanding, directly connecting the organic interests of "Pinners" with highly relevant, personalized content.
What makes this position unique is the scale of the visual and graph-based data. You will not only build deep learning architectures but also design the end-to-end data pipelines that feed them. This requires a tight integration of software engineering discipline, big data infrastructure, and advanced machine learning theory.
Recruiter Call
reportedHalf of this call is the part candidates treat as small talk: start date, notice period, work authorisation and its timing, location and time zone, on-call, and the number. Those are what kill offers late, after several engineers have each spent a day. Surfacing a hard constraint now costs you nothing and occasionally buys you something, since a loop compressed to fit a competing deadline can usually only be arranged if it is asked for early. The common failure is deflecting the compensation question twice, then discovering at offer stage that the band never reached your number.
What to demonstrate
- Whether your hard constraints are compatible with the role before a loop gets booked: earliest start, notice period, what authorisation you hold and when it needs action, days on site, willingness to carry a pager
- Whether you give a compensation range with something behind it, such as current total compensation or a competing timeline, rather than leaving the band untested
- Whether your stated timeline is real, since a competing deadline raised now is something scheduling can sometimes work around and the same deadline raised at offer stage usually is not
How to prepare
- Write each constraint down in one line before the call and state them as facts rather than negotiating them live under a question you were not expecting
- Set your range from two or three current data points for that level and location, and name the structure you are quoting in, so the number is comparable to the one they are holding
- If another process is running, say where it stands and by when, and ask directly whether this loop can be scheduled inside that window
Technical Phone Screen
reportedHalf of this call is the part candidates treat as small talk: start date, notice period, work authorisation and its timing, location and time zone, on-call, and the number. Those are what kill offers late, after several engineers have each spent a day. Surfacing a hard constraint now costs you nothing and occasionally buys you something, since a loop compressed to fit a competing deadline can usually only be arranged if it is asked for early. The common failure is deflecting the compensation question twice, then discovering at offer stage that the band never reached your number.
What to demonstrate
- Whether your hard constraints are compatible with the role before a loop gets booked: earliest start, notice period, what authorisation you hold and when it needs action, days on site, willingness to carry a pager
- Whether you give a compensation range with something behind it, such as current total compensation or a competing timeline, rather than leaving the band untested
- Whether your stated timeline is real, since a competing deadline raised now is something scheduling can sometimes work around and the same deadline raised at offer stage usually is not
How to prepare
- Write each constraint down in one line before the call and state them as facts rather than negotiating them live under a question you were not expecting
- Set your range from two or three current data points for that level and location, and name the structure you are quoting in, so the number is comparable to the one they are holding
- If another process is running, say where it stands and by when, and ask directly whether this loop can be scheduled inside that window
Virtual Onsite Loop
reportedA day like this is several different games in a row, and the expensive mistake is carrying the previous one into the next room. Coding rewards narrow precision and finishing inside a timer. Design rewards breadth, stated assumptions and naming what you are deliberately not building. Behavioural rewards specificity about people and decisions. Candidates who over-engineer a coding problem they were supposed to finish, or who start sketching class hierarchies before anyone has agreed what the system has to do, are usually still playing the last round. Between rooms, name out loud which game the next one is.
What to demonstrate
- Whether the coding round ends with something that runs and has been traced against a degenerate input, rather than an extensible design that was never finished
- Whether a design discussion opens by agreeing on traffic shape, read-to-write ratio and what is allowed to be stale, instead of proceeding from an architecture you arrived with
- Whether a behavioural answer names a person, a disagreement and what you did about it, rather than describing the system the story happened inside
- Whether the opening habits still appear late in the day: restating the problem, asking for constraints, saying the plan before typing
How to prepare
- Book three mocks of different types back to back on one afternoon and ask each interviewer afterwards which round you answered in the wrong mode
- Write a three-line opening script per round type — coding: restate, name the approach and its cost, then type; design: ask for scale, read-write mix and what must not break; behavioural: name the person, the stakes and the decision — and run it off a card so the switch is mechanical rather than remembered
- Practise coding with a timer you do not extend, stopping when it stops, so the trained reflex is to finish a correct solution rather than to keep improving one
4 candidate reports. Individual accounts describe a particular role and hiring cycle.
Pinterest Machine Learning Engineer interview: ML theory and quick rejection
The loop was short but rough. In the initial conversation, I tried to explain my projects, but the interviewer did not seem willing to listen. I then had a technical machine-learning session covering core theory, including vanishing gradients and ensemble methods such as bagging and boosting. There was also a LeetCode-style problem for about 30 minutes. The process ended quickly, and I received a…
Read full experiencePinterest Senior+ Machine Learning Engineer Interview Experience — Five Onsite Rounds, an Add-On Round, and a Scheduled Rejection Email
View report detailsPinterest Senior+ Machine Learning Engineer Interview Experience — Two ML Design Rounds, Two Buggy Codings, and a Gut Feeling of Rejection
View report detailsPinterest Intern Machine Learning Engineer Interview Experience — 70-Minute CodeSignal OA with Hand-Written Bagging and Naive Bayes
Based in Toronto. Looks like several teams were hiring at the same time, mostly on the search/ads/recommendations side. HR reached out on LinkedIn, and after I submitted my resume and cover letter they sent me an OA. 70 minutes, 10 questions, done on CodeSignal — 6 multiple choice, 1 coding question, and 2 ML engineering questions. The multiple choice questions were mostly ML fundamentals: the de…
Read full experiencePracHub editorial advice for the preparation topics above.
Paginating with LIMIT/OFFSET over a set that changes while the client is reading it
OFFSET n makes the database produce and discard n rows before returning anything, so the cost of a page grows with its depth rather than with its size and page 500 costs five hundred pages of work. The correctness problem is worse than the cost: if a row is inserted or reordered between two page fetches, rows shift across the offset boundary and are either skipped entirely or returned twice, and neither outcome leaves any trace in the response for the client to detect. Keyset pagination - WHERE (sort_key, id) < ($last_sort_key, $last_id) ORDER BY sort_key DESC, id DESC LIMIT n, backed by an index in exactly that order - reads only the rows it returns and is stable against concurrent inserts. It requires the tie-break column: a timestamp is not unique, and duplicate sort keys straddling a page boundary reintroduce the skip it was adopted to remove.
Letting a slow dependency consume unbounded concurrency
The failure that takes a service down is usually not an error but a delay. A dependency answering in thirty seconds instead of fifty milliseconds holds each request's worker or connection six hundred times longer, and since required concurrency is arrival rate times latency, a fleet sized for sixty in-flight requests now needs thirty-six thousand to sustain the same rate - so it queues, and requests whose clients have already abandoned them still occupy resources. Retries make it precisely worse: a policy of three attempts triples the load on a dependency at the exact moment it is least able to serve, which is how one slow dependency becomes an outage of everything sharing that pool. Containment is four specific things - a timeout on every outbound call shorter than the caller's remaining budget, a bounded pool per dependency so one cannot starve the others, backoff with full jitter rather than a fixed delay so retries do not resynchronise, and a circuit that stops sending once the failure rate makes an attempt pointless.
Optimising an axis nobody named
Ask which resource is actually scarce here: wall-clock latency, throughput, memory footprint, cost per request, or engineering time. Shaving a constant factor off an in-memory step is wasted effort when the same function makes a blocking remote call inside the loop.
Finishing a solution without stating its complexity
Give time and space in the same breath as the code, and define n explicitly when there are two sizes, since n nodes and m edges are not interchangeable. Space is the half that gets skipped: count the auxiliary structures you allocate and the recursion stack at its deepest, not only the answer you hand back.
Choose a category, try a prompt, then open its approach, worked solution or follow-up when you need it.
Near-Duplicate Clustering: Given a mapping from every image ID to a li…
Near-Duplicate Clustering: Given a mapping from every image ID to a list of its near-duplicate image IDs (e.g., "A": ["B", "I", "K"]), write an algorithm using BFS, DFS, or Union-Find to group these images into distinct, connected near-duplicate clusters.
Approach
- Pick the metric from the cost of each error type, not from habit.
- Name the simplest model that could work and what would make you move past it.
- Say how you would validate it, and where leakage could enter the split.
Follow-up
- What changes if the classes are heavily imbalanced?
- Where could label leakage enter this setup?
Vanishing Gradient Problem: What causes gradients to vanish in deep ne…
Vanishing Gradient Problem: What causes gradients to vanish in deep neural networks? Discuss how different activation functions (such as ReLU, Leaky ReLU, and ELU) mitigate this issue, and explain their mathematical trade-offs.
Approach
- State the learning problem: the label, the unit of prediction and how the model is used.
- Pick the metric from the cost of each error type, not from habit.
- Say how you would validate it, and where leakage could enter the split.
Follow-up
- What changes if the classes are heavily imbalanced?
- Where could label leakage enter this setup?
L1 vs. L2 Regularization: Explain the mathematical difference between …
L1 vs. L2 Regularization: Explain the mathematical difference between L1 and L2 regularization. Be ready to write out the optimization objectives and derive why L1 regularization mathematically leads to feature sparsity while L2 regularization leads to weight decay.
Approach
- State the learning problem: the label, the unit of prediction and how the model is used.
- Pick the metric from the cost of each error type, not from habit.
- Say how you would validate it, and where leakage could enter the split.
Follow-up
- Where could label leakage enter this setup?
- What changes if the classes are heavily imbalanced?
Fully Connected Networks: Explain the forward and backward propagation…
Fully Connected Networks: Explain the forward and backward propagation steps in a fully connected layer. How do you optimize memory consumption during backpropagation?
Approach
- Say how you would validate it, and where leakage could enter the split.
- State the learning problem: the label, the unit of prediction and how the model is used.
- Pick the metric from the cost of each error type, not from habit.
Follow-up
- How would you know the model is overfitting?
- Where could label leakage enter this setup?
Sparse Matrix Representation: Implement a memory-efficient sparse matr…
Sparse Matrix Representation: Implement a memory-efficient sparse matrix class from scratch without using any external linear algebra libraries. Implement custom methods for matrix addition and matrix multiplication that handle different sparsity patterns efficiently.
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- Name the brute-force solution and its complexity before improving on it.
- Walk one small example through your approach before writing the whole thing.
Follow-up
- Which test case would catch an off-by-one here?
- How does this change if the input no longer fits in memory?
Multi-Dimensional BFS: Implement a multi-dimensional breadth-first sea…
Multi-Dimensional BFS: Implement a multi-dimensional breadth-first search to traverse complex grid structures or dependency graphs under specific routing constraints.
Approach
- Restate the input: its shape, its size, and what is guaranteed about it.
- State the target complexity and say which constraint rules the naive version out.
- Walk one small example through your approach before writing the whole thing.
Follow-up
- Which test case would catch an off-by-one here?
- What is the worst case, and how likely is it on real data?
Diff a projection against the primary without per-row point reads
The listing projection has drifted and some rows show a stale version. The primary holds 40,000,000 resource rows across 12,000 tenants while serving 1,200 writes and 14,000 reads per second. The obvious repair, reading each resource row and comparing its version against the projection, is correct and would eventually finish. Explain precisely why it is unacceptable here, then give a diff that finds the differing rows, state its complexity, and make it safe to run against a live primary. Replication lag is usually under 100 ms and is not bounded.
Approach
- Quantify the naive cost rather than calling it slow: 40,000,000 point reads at even 0.5 ms each is over five hours serialised, and the only lever is concurrency, which is exactly what you cannot spend. The primary's pool is sized for the write path, and 40,000,000 random reads evict the buffer cache that sustains the 85 percent cache hit rate, so the audit degrades the system it is auditing.
- Replace random access with one ordered pass per side. Both sides can be read in (tenant_id, resource_id) order, which is a sequential scan on each and a merge join in O(n) time and O(1) memory. For a dense diff that is the whole answer, and it reads the primary once instead of 40,000,000 times.
- For the expected sparse case, compare range hashes instead of rows: partition the key space, compute per range an order-independent aggregate over hash(resource_id, version), compare aggregates, and descend only into ranges that differ. With d differing rows and branching factor B, at most d ranges mismatch per level, so the drill-down examines O(d log_B(n/d)) ranges and reads full rows only in mismatching leaves.
- Aggregate with a sum modulo 2^64 or a multiset hash, never XOR. XOR is order-independent but self-cancelling, so two rows wrong in the same way, or a row duplicated on one side, leave the range aggregate matching and the range is declared clean.
- Pin the comparison to a point in time or it reports lag as drift: consider only rows whose updated_at is older than now minus a lag margin, and re-check each candidate mismatch individually before repairing. At 1,200 writes per second a diff without this reports thousands of false positives, and an unattended repairer would then overwrite live rows with stale values.
- Make the run resumable and throttled: batch by range key, persist the last completed range, and watch a signal such as replica lag or primary CPU, pausing rather than pressing on. A reconciliation that cannot be stopped and resumed gets killed halfway and restarted from zero, which is how a repair becomes an incident.
Worked solution 35 min
- Compute the naive cost explicitly at 40,000,000 reads and 0.5 ms each, then at 100 concurrent, and state what those connections do to a pool already carrying 1,200 writes per second.
- Write the merge-join version over (tenant_id, resource_id) and state its memory.
- Define the range aggregate: the range key, the per-row hash input, and the combining function, with one sentence excluding XOR.
- Work an example with 40,000,000 rows, branching factor 256 and 5 differing rows, and count the ranges examined.
- Add the watermark filter and the resume point, and name the throttle signal the loop watches.
Follow-up
- The diff reports 900 stale rows. How do you decide between patching those rows and rebuilding the projection from resource_revision?
- Same job, but the projection lives in a search index that cannot be scanned in key order. What changes?
- How would you run this continuously at low cost instead of only as incident response?
Keep soft-deleted accounts from blocking re-registration
app_user holds user_id, tenant_id, email CITEXT, password_hash (NULL for SSO principals), email_verified_at, auth_version, status ('invited','active','suspended','deactivated'), created_at, updated_at, deleted_at. Two live accounts for one address inside a tenant must be impossible, but an address freed by a soft delete must be reusable, and the same tenant may delete and re-register it repeatedly. Write the uniqueness DDL for PostgreSQL 16, then the equivalent for MySQL 8 where partial indexes do not exist, and say what each permits once three deleted rows already hold that address.
Approach
- Start from what is actually unique: not (tenant_id, email), but (tenant_id, email) among live rows. PostgreSQL says that directly — CREATE UNIQUE INDEX app_user_live_email ON app_user (tenant_id, email) WHERE deleted_at IS NULL. A full constraint over the same two columns burns the address permanently the first time someone deletes an account.
- Keep case-insensitivity in the type or the index, never in the application: CITEXT as given, or UNIQUE (tenant_id, lower(email)) as an expression index where the extension is unavailable. A case-sensitive unique column is exactly how two accounts for one human appear.
- For MySQL 8 the predicate has to move inside the key: add a discriminator column that is a constant 0 while the row is live and is set to user_id on delete, with UNIQUE (tenant_id, email, deleted_marker). Live rows share the constant and still collide; deleted rows differ from each other and stop colliding.
- State the NULL variant and its dependency: leaving the marker NULL for deleted rows also works, because a unique index treats NULLs as distinct — true in MySQL, and true in PostgreSQL only under the default NULLS DISTINCT, which PostgreSQL 15 lets you reverse. Check the polarity against the three existing deleted rows: constant-on-live is what preserves the collision you want, and reversing it silently admits duplicate live accounts.
- Say what a soft delete must do besides setting deleted_at: increment auth_version so existing tokens stop validating, leave resource.owner_user_id and resource_revision.actor_user_id intact, and accept that the address is retained — erasure is a different requirement answered by scrubbing the column, not by a DELETE that would break those references.
Follow-up
- A deleted account re-registers with the same address the next day. Do the old resource rows follow the new user_id, and how does the API keep the two principals apart?
- How do you honour an erasure request while resource_revision.actor_user_id still references this table?
- What changes if a user may hold membership in two tenants?
Hold a per-tenant active cap against concurrent creates
A tenant on the standard plan may hold at most 50 resources with status='active'. The create handler runs SELECT count(*) FROM resource WHERE tenant_id = $1 AND status = 'active', compares to 50, then inserts. Two creates arrive 3 ms apart on different instances and the tenant lands at 51. Name the anomaly, say whether PostgreSQL 16 READ COMMITTED or REPEATABLE READ prevents it and why, then give an implementation that holds the cap at READ COMMITTED with the exact statements. Finally, say what changes when the cap is 'at most one running export per tenant' on job_run.
Approach
- Name it: write skew. The two transactions read an overlapping set and write disjoint rows, so there is no row-level conflict for the engine to detect and each commit is individually legal.
- Rule out the levels precisely. READ COMMITTED takes a fresh snapshot per statement and takes no lock on the counted rows, so both see 49. PostgreSQL's REPEATABLE READ is snapshot isolation: it removes non-repeatable reads and phantoms within the snapshot but still admits write skew, because the anomaly is not a re-read of a changed row, it is a read of a set that a concurrent transaction invalidates. Only SERIALIZABLE closes it, by tracking the read dependency and aborting one transaction with SQLSTATE 40001 — a guarantee that exists only if the application re-runs the whole transaction from the read.
- Convert the set predicate into a single-row conflict: keep tenant.active_resource_count and run UPDATE tenant SET active_resource_count = active_resource_count + 1 WHERE tenant_id = $1 AND active_resource_count < 50 in the same transaction as the INSERT. Zero affected rows is the cap, returned as 409. The row lock serialises the decision at any isolation level, and contention is bounded to one tenant's row — which is also the fair-scheduling unit, unlike a global counter that would convoy every tenant behind one row.
- State the cost you just took on: a counter is a second source of truth that can drift, so every path that changes status must adjust it inside the same transaction, and a periodic reconciliation has to exist, with resource_revision as the authority for what the count should have been.
- For the job case the invariant is expressible per row, so let the database hold it: a partial unique index on job_run (tenant_id, job_type) WHERE status IN ('queued','running') makes a second running export unwritable and the loser takes 23505, mapped to 409. That is strictly better than a counter — no drift, no reconciliation — and it is available only because the cap is one rather than fifty.
- Add the retry discipline each route demands: under SERIALIZABLE both 40001 and deadlock 40P01 are retryable and the retry must re-execute the read, while under READ COMMITTED with the counter nothing retries, because the conflict is reported to the caller rather than raised as an error.
Worked solution 35 min
- Reproduce with two sessions that both count 49, both insert and both commit, at READ COMMITTED and then at REPEATABLE READ; record the final active count for each.
- Repeat both sessions at SERIALIZABLE and record which SQLSTATE the loser receives and at which statement it is raised.
- Implement the counter form and run a 20-way concurrent create against a tenant sitting at 45 active resources.
- Implement the partial unique index for the job case and race 20 enqueues of the same export.
Follow-up
- A resource moves from archived back to active. Which statements change, and what breaks if the counter update and the status change land in different transactions?
- The cap becomes plan-dependent and a plan can change mid-month. Where does the number 50 live, and who reads it?
- How do you detect after the fact that the counter drifted, without locking the table?
Fraud and Spam Detection: Design a real-time system to detect fraudule…
Fraud and Spam Detection: Design a real-time system to detect fraudulent accounts or spam Pins. Address how you would handle adversarial drift and latency constraints.
Approach
- Fix the product goal and the online metric before choosing any model.
- Say where features come from at serving time and how they match training.
- Name what you would monitor after launch and what triggers a retrain.
Follow-up
- How would you roll the new model out safely?
- What happens when a feature is missing at serving time?
Pinterest Homepage Recommendation: Design the recommendation engine fo…
Pinterest Homepage Recommendation: Design the recommendation engine for the Pinterest Homefeed. Explain your strategy for candidate retrieval, heavy ranking, scoring, and diversity filtering.
Approach
- Separate the offline training path from the online serving path.
- Fix the product goal and the online metric before choosing any model.
- Name what you would monitor after launch and what triggers a retrain.
Follow-up
- How would you roll the new model out safely?
- What happens when a feature is missing at serving time?
Make resource creation safe for a client that retried a timeout
POST /v1/resources creates a resource row and, in the same transaction, a resource_revision and an outbox_event. A partner's HTTP client times out at 2 seconds and retries twice; under load the first request is still executing when the retry arrives, and duplicate resources are appearing. Using idempotency_key (tenant_id, idempotency_key, request_fingerprint, state, response_status, response_body, resource_id, locked_until, expires_at), specify the whole contract: first call, concurrent retry, retry after completion, a key reused with a different body, and what the client does with each response.
Approach
- Name why the plain REST shape fails here. POST is not idempotent, a timeout leaves the client unable to distinguish a lost request from a lost response, so it must retry - and the retry is a second create. Checking whether the key exists and then inserting it is check-then-act: under exactly the concurrency that caused the timeout, both attempts look, both see nothing, both proceed.
- Let the unique constraint arbitrate instead. INSERT the key row with state 'in_flight' and locked_until set to now plus the request deadline, and commit that reservation on its own before any work is attempted - an uncommitted row is invisible to the concurrent retry, which is the entire point of writing it first. One caller wins the insert and proceeds; the other takes the unique violation and reads the row rather than falling through to the work.
- Branch the loser on what it found. 'succeeded' replays response_status and response_body verbatim, with a header marking it a replay. 'in_flight' with locked_until in the future returns 409 and a Retry-After - do not park a request thread waiting on another request, that is how a concurrency bound is consumed by retries. 'in_flight' past locked_until is reclaimable by a conditional UPDATE predicated on the old locked_until, so exactly one reclaimer wins. A fingerprint mismatch is 422 and never the stored response: a different body under the same key is a client defect, and serving the old response hides it forever.
- Commit the effect and the key's completion together - resource insert, revision row, outbox row, and the UPDATE of the reserved key to 'succeeded' with the stored response, in one transaction. That makes a crash binary: either all of it is durable and the key reads 'succeeded', or the transaction rolls back and the only durable trace is the reservation row, still 'in_flight', with no resource behind it. The reclaim branch finishes that case - once locked_until passes, one retry takes the row by conditional UPDATE and runs the work for the first time. Splitting the two commits is what creates the window with no recovery: the effect durable, the key still 'in_flight', and a reclaimer that cannot distinguish that from work never started, so it creates a second resource.
- State the lifetime explicitly. Keys expire - 24 hours is a common choice - and are swept, so the client's total retry window must be shorter than the TTL, or a late retry re-executes with no record that it is a duplicate. The key is scoped by tenant so two tenants choosing the same string do not collide, and the client generates one key per logical attempt, not per network send.
- Write the client rules: on timeout, retry the same key with capped backoff; on 409, wait Retry-After and retry the same key; on 422, stop and fix the caller; on 2xx, treat the body as authoritative whether it was executed or replayed.
Worked solution 30 min
- Write the two-request interleaving that defeats read-then-insert, with the line each request is executing at each step.
- Write the insert-first sequence and the four branches taken on unique violation, as a table of state, locked_until and response.
- Put the completion UPDATE inside the same transaction as the resource, revision and outbox writes, then state what a crash before that commit leaves durable and which branch cleans it up.
- Define the key's scope, lifetime and generation rule, then bound the client's retry budget by the TTL.
- Fire two concurrent requests carrying one key and count rows in resource and resource_revision.
Follow-up
- Response bodies are pruned at 24 hours but resource_id is kept. What do you return to a replay that arrives after pruning?
- The endpoint now also sends a welcome email. Which part of this contract prevents a second send, and which part cannot?
- What exactly breaks if the key is scoped globally rather than per tenant?
p99 jumped on one listing filter while p50 stayed flat
After a release that added an owner_user_id filter to the resource listing, p99 rose from 90 ms to 1.9 s while p50 stayed at 40 ms. Traffic and row counts are unchanged. resource carries the index (tenant_id, status, updated_at DESC, resource_id DESC). The new query filters tenant_id and owner_user_id, orders by updated_at DESC, resource_id DESC, and takes 20 rows. On PostgreSQL, explain the shape of the regression, prove it from a query plan, and give the index you would add.
Approach
- Start from the shape. A flat p50 with a moved p99 means a subset of requests changed cost, not all of them, so the first job is naming the subset. Bucket the endpoint's latency by the tenant's row count; the natural hypothesis is that large tenants are a small share of requests and all of the tail.
- Get the plan for the new query on a large tenant with EXPLAIN (ANALYZE, BUFFERS). Expect an index scan over the tenant's range, a filter discarding most of it, then a Sort feeding the Limit, possibly reporting Sort Method: external merge Disk. Read actual rows on the scan node, not estimated.
- Explain why the existing index cannot serve it. A composite B-tree is seekable only as a left prefix, and with no equality predicate on status the scan cannot treat updated_at as an ordering, because rows in the tenant's range are ordered by status first. Everything matching must be read and sorted before LIMIT 20 can apply, so a tenant with 400,000 rows pays 400,000 rows to return 20.
- Add (tenant_id, owner_user_id, updated_at DESC, resource_id DESC). Equality on the first two columns leaves the index ordered by updated_at within that pair, so the plan becomes an index scan that stops after 20 rows with no Sort node. PostgreSQL can scan a B-tree backwards, so the DESC markers matter only if the two sort columns ever disagree in direction; keeping them explicit documents the order the keyset cursor depends on.
- Price the fix. This is a fourth index on a table taking about 1.2k writes/second, and every insert and version bump maintains it. Justify it against the query it serves, and check whether it makes an existing index redundant, which here it does not, since the original still serves the status-filtered default listing.
- Re-measure per tenant-size bucket rather than in aggregate. A fleet-wide p99 can improve while the largest tenant is still on the old plan.
Follow-up
- The endpoint paginates with OFFSET. What does page 500 cost with your index, and what does the keyset version cost?
- How would you have caught this before release, given that a 10,000-row seed database produces the same plan shape at an unnoticeable cost?
- If a fourth index were unacceptable on write grounds, what else could serve this query?
Four days sample coding, design, fundamentals and the practical rounds at deliberately shallow depth, which is enough to surface the topics you did not know were in scope. That map, rather than a guess made on day one, decides where the last three days go.
Prepare, practise & reflect
One practical outcome each day. Spend longer where you need it.
0 / 7 done01Coding, one pass at shallow depth
- Solve one problem from each of six families, an array with two pointers, hash counting, binary search, a tree traversal, a graph traversal and one dynamic program, under a hard twenty-minute cap with no extensions, marking each finished, late, or stalled.
- For every stall, write the exact move you could not make rather than the subject, so the note reads could not turn the recurrence into a loop rather than bad at dynamic programming.
- Fix nothing today. The value of the pass is the unfixed record.
Deliverable: Six timed attempts marked finished, late or stalled, each stall carrying a named blocking move.
Practice prompt ↗Practice prompt ↗Practice prompt ↗Worked solution ↗02Design, one pass at shallow depth
- Spend twenty minutes each on three different shapes, a read-heavy feed, a write-heavy ingest path, and something needing a transaction across two entities, stopping each at requirements, interface and data model.
- After each, write the first question you could not answer, which is usually a number you could not estimate or a failure mode you had no vocabulary for.
- Mark which of the three you would be most relieved not to be asked, and treat that as data rather than as a preference.
Deliverable: Three shallow designs, each with the first unanswerable question written at the bottom.
Practice prompt ↗Practice prompt ↗Practice prompt ↗03Fundamentals and the practical rounds
- Answer eight short questions in writing at four minutes each, covering the material that fills the gaps between the big rounds: what happens between a URL and a rendered page, what an index costs on write, when a process is preferable to a thread, and what conditions a deadlock requires.
- Do one thirty-minute practical task of the kind a take-home compresses: read an unfamiliar two-hundred-line file and write what it does, what you would change, and the one thing you remain unsure of.
- Score every answer fluent, correct but slow, or absent, and keep the absent ones visible.
Deliverable: Eight scored short answers and one written reading of unfamiliar code.
Practice prompt ↗Practice prompt ↗04The rounds that are about you, and the map
- Deliver three behavioural answers aloud against a timer, a conflict, a failure you owned, and a decision made without enough information, marking any that ran past three minutes or contained no number.
- Assemble the map: every marked item from days one to three on a single page, sorted by how likely it is to appear in your loop rather than by how uncomfortable it felt.
- Choose exactly two areas for the remaining three days and write down what you are deliberately abandoning.
Deliverable: A one-page scored map of the whole surface area with two areas chosen and the rest explicitly abandoned.
Practice prompt ↗Practice prompt ↗Worked solution ↗05First chosen area, to the depth you skipped
- Work the higher-ranked area in four focused blocks, choosing items one level above where you stalled rather than repeating what already works.
- After each block write the rule you extracted in one sentence with its precondition attached, since a rule carrying no precondition is exactly what fails under a variation.
- Re-attempt the day-one or day-two item that exposed this area and compare against the original timing.
Deliverable: Four worked blocks, a timed re-attempt against the original, and three one-sentence rules with preconditions.
Practice prompt ↗Practice prompt ↗06Second chosen area, where the gap is coverage rather than speed
- Treat the second area differently from the first. Day five drilled something you could already half-do; this one is usually a topic you had simply never met, so build one worked reference example end to end and keep it, rather than attempting six problems badly.
- Write down the vocabulary you were missing on day two or three, five terms at most, each with the one sentence that makes it usable in an answer rather than the textbook definition.
- Redo the shallow attempt that exposed this area and note whether you now fail later in the problem, because moving the failure point is the realistic gain from a single day and is worth more than a score that did not change.
Deliverable: One worked reference example for the newly covered area, a five-term vocabulary list, and a note on where the failure point moved.
Practice prompt ↗Practice prompt ↗07Reassemble the loop
- Sit two rounds back to back with no gap, ordering them so the area you chose second comes last, because the map was built from rested, isolated attempts and the loop will reach your weaker area when you are already spent.
- Write where the second round suffered from the first, which is normally the point at which structure collapses into narration.
- Reduce the week to one page holding only the rules you can state without reading them.
Deliverable: Mock notes on cross-round carryover plus a one-page card of rules you can recite from memory.
Practice prompt ↗Practice prompt ↗Worked solution ↗Expand any day for tasks and deliverables. Your progress is saved on this device.
For anything that touched live traffic, be ready to say how you would have undone it: a flag, a staged rollout, dual writes with the old path still authoritative. Once the old column is dropped or the source rows are overwritten there is no reverse, so name what you kept a copy of and for how long.
Argue against a design, lose, and commit anyway
Describe a design you argued against and lost. State the failure you predicted as a named mechanism, not a feeling about complexity: two services that would need one transaction, a projection with no rebuild path, a write path with no idempotency key. Say what evidence you brought, what the decision maker weighed instead, and what you did after the decision was made: what you instrumented, what you wrote down, and whether the prediction came true. Five minutes.
Approach
- State the prediction in falsifiable form up front: the mechanism, the condition that triggers it, and the observable outcome. A prediction that cannot be checked also cannot be credited to you later.
- Show the evidence you had at the time and label each piece honestly as measured, analogous, or intuition. Keeping the intuition is fine; disguising it as data is the thing that erodes your standing in the next argument.
- Represent the opposing case at full strength, including the constraint you did not control: a fixed date, a team boundary, or the fact that the decision was cheap to reverse and yours was not.
- Make disagree-and-commit concrete. Name the artefact you left behind so the prediction could be settled without you: the alert and its threshold, the counter on the dashboard, the decision note that recorded the trade-off and the condition that would revisit it.
- Report the outcome without editing it. If the design held and your predicted mechanism never fired, say so and say what you had mis-weighted, which is more persuasive than a vindication story.
Follow-up
- What threshold on that alert would have proved you right, and did anyone ever look at it?
- If the same proposal arrived tomorrow with the same deadline, would you argue it the same way?
- How did you behave toward the design once it shipped and started failing in a different way than you predicted?
Narrate an outage you owned from page to postmortem
Pick an incident you personally drove, ideally one where writes were affected rather than reads. In six to eight minutes: state the symptom as it first appeared on a dashboard, the blast radius you established before you knew the cause, the mitigation you applied and when, the mechanism you eventually proved, and the follow-up that would prevent a repeat. Bring numbers: error rate, tenants affected, minutes to mitigate, minutes to resolve. If you cannot name what you measured, choose a different incident.
Approach
- Open on the signal rather than the cause: which metric at which percentile moved, on which service, at what time, so the listener follows the same evidence you had rather than a conclusion you already reached.
- Separate mitigation from diagnosis out loud. State what you did to stop the bleeding (flag off, shed traffic, drain a lease, roll back a deploy) and say plainly that you did it before the mechanism was known, because those are two jobs with different deadlines.
- Establish blast radius in countable terms: how many tenants, how many writes, and crucially whether the effect was loss or only delay. An append-only revision table or a pending outbox row means the change survived and the projection was merely behind, which is a repair rather than a data-loss incident.
- Prove the mechanism instead of asserting it. Name the trace span that grew, the plan that flipped to a sequential scan, the lease that expired, plus one alternative you ruled out and the signal that stayed flat while you ruled it out.
- Close on the durable fix and its cost, distinguishing what landed that week from what needed an expand-and-contract migration across several deploys, and say which of the two you actually finished.
Follow-up
- What would you do differently in the first five minutes, given the same dashboard and no more information?
- Which follow-up action did you deliberately not take, and why was dropping it the right call?
- How did you convince yourself the mitigation was safe to apply while the cause was still unknown?
Tell callers you do not own that their integration breaks
A field in a write endpoint's response must change shape. You own the endpoint; you do not own the four internal callers or the outbound webhook consumers who read it. Describe a deprecation you were responsible for: what you shipped first, how you established who was actually reading the field, the window you gave and what set its length, what you did about the consumer who never moved, and how you decided removal was safe. Name the signal you used, not the announcement you sent.
Approach
- Establish the reader set empirically rather than from a wiki of owners: per-field usage counters keyed by principal, or access logs attributed to a consumer. State the blind spot of whichever you pick, since a consumer that reads the field only on a monthly job will not appear in a week of logs.
- Ship additive first. Populate the new field alongside the old one so no reader is forced to move, which is also what keeps a rolling deploy safe, because old and new instances answer the same requests at the same time and a rollback must still find the old shape present.
- Set the window from the slowest legitimate consumer's release cadence, not from your calendar, and decide separately what to do for a consumer with no release process at all, such as an external webhook endpoint you can only email.
- Convert silence into evidence before you rely on it: a short, low-traffic removal window that makes a still-dependent consumer fail visibly and loudly while you are watching, rather than at three in the morning after you have moved on.
- State the removal criterion as a measurement with a duration attached, such as observed reads at zero across a full billing cycle, and keep the change reversible for one release after removal.
Follow-up
- How would you detect a consumer that reads the field only during a monthly export?
- One caller refuses to move and has a commercial relationship behind it. What changes in your plan and what does not?
- After removal, what makes the change irreversible, and how long before you cross that line?
- 01
Describe a design you argued against and lost. State the failure you predicted as a named mechanism, not a feeling about complexity: two services that would need one transaction, a projection with no rebuild path, a write path with no idempotency key. Say what evidence you brought, what the decision maker weighed instead, and what you did after the decision was made: what you instrumented, what you wrote down, and whether the prediction came true. Five minutes.
- 02
Pick an incident you personally drove, ideally one where writes were affected rather than reads. In six to eight minutes: state the symptom as it first appeared on a dashboard, the blast radius you established before you knew the cause, the mitigation you applied and when, the mechanism you eventually proved, and the follow-up that would prevent a repeat. Bring numbers: error rate, tenants affected, minutes to mitigate, minutes to resolve. If you cannot name what you measured, choose a different incident.
- 03
A field in a write endpoint's response must change shape. You own the endpoint; you do not own the four internal callers or the outbound webhook consumers who read it. Describe a deprecation you were responsible for: what you shipped first, how you established who was actually reading the field, the window you gave and what set its length, what you did about the consumer who never moved, and how you decided removal was safe. Name the signal you used, not the announcement you sent.
Is this an official Pinterest interview guide?
No. It is PracHub's own research and practice material for the Machine Learning Engineer role at Pinterest. Rounds and questions reflect what candidates have reported, not a process Pinterest has published, and they change over time. Confirm the current format and scope with your recruiter.
PracHub interview research ↗How difficult are the coding rounds compared to other top-tier tech companies?
Candidates frequently report that the coding rounds are challenging, often equivalent to medium-to-hard algorithmic problems. The primary difficulty lies in the strict expectation for bug-free code, optimal complexity, and the requirement to handle complex edge cases within a 45-minute window.
PracHub interview research ↗What is the online assessment (OA) like?
The online assessment is highly time-constrained. You will typically be given 70 minutes to complete a mix of multiple-choice questions on machine learning theory and several coding problems. Speed is critical; many candidates fail because they run out of time, so practicing rapid execution is essential.
PracHub interview research ↗How deeply does Pinterest test machine learning math?
Very deeply. Unlike companies that only focus on high-level system design, Pinterest expect you to know the underlying mathematics of machine learning algorithms. You should be prepared to write down loss functions, derive gradients, and explain the mathematical differences between optimization techniques on a whiteboard.
PracHub interview research ↗What is the hybrid work policy at Pinterest?
Pinterest operates under a flexible working model called "PinFlex." This model allows employees to work from home, in an office, or a combination of both, depending on the team's requirements and the candidate's location.
PracHub interview research ↗Sources & methodology 3 sources ↗
Official role evidence, timestamped platform data and clearly labeled preparation advice.
- 01PracHub interview research ↗
PracHub editorial research into this company and role, maintained with this guide. Candidate-reported, not an employer publication.
platform · Accessed 2026-09-30 - 02PracHub Machine Learning Engineer practice ↗
Cross-company practice questions for this role.
platform · Accessed 2026-09-30 - 03PracHub interview preparation framework ↗
The framework the preparation plan follows.
platform · Accessed 2026-09-30
Explore related engineering environments and compare the preparation focus.
Snap Machine Learning Engineer Interview Questions & Guide 2026
What Snap asks Machine Learning Engineer candidates, from reported rounds and questions.
Reddit Machine Learning Engineer Interview Questions & Guide 2026
What Reddit asks Machine Learning Engineer candidates, from reported rounds and questions.
Nextdoor Machine Learning Engineer Interview Questions & Guide 2026
What Nextdoor asks Machine Learning Engineer candidates, from reported rounds and questions.
LinkedIn Machine Learning Engineer Interview Questions & Guide 2026
What LinkedIn asks Machine Learning Engineer candidates, from reported rounds and questions.