← Back to Directory

Pinterest interview notes, page 22 of 24

Large Enterprises

Most of these Pinterest write-ups come from Machine Learning Engineers (91), then Software Engineers (77) and Data Scientists (51). The most reported rounds are the technical phone screen (168), the system design round (37), the product sense round (14) and the behavioral / leadership round (5).

235 interview notes · updated Sep 2026

Pinterest·Machine Learning Engineer·Technical Phone Screen

Jul 2026
Pinterest ML Engineer interview with a tricky warehouse packing problem that adds a twist to a classic LC variant. The bidirectional push mechanic is what makes it interesting and also where I got tripped up.
  • You have two arrays representing box heights and warehouse room heights. Boxes can be pushed in from either the left or the right side, and each box must fit within the effective minimum height along its entire push path. What is the maximum number of boxes you can fit in the warehouse?

“The bidirectional part is what got me.”

View Post

Pinterest·Machine Learning Engineer·Onsite - System Design / Architecture

Jul 2026
Pinterest ML engineer interview focused entirely on a large-scale URL health detection system. One long design question that touched basically every layer of the stack, from crawling strategy to downstream ad system integration. Dense but interesting.
  • Design an ML system to detect broken or low-quality landing pages (404s, soft-404s, parked domains, redirect loops) across a massive URL inventory. Walk through labeling, features, crawling strategy, model architecture, serving, and success metrics.

“This question is basically six questions stitched together and they want you to move through all of them without getting stuck.”

View Post

Pinterest·Machine Learning Engineer·Technical Phone Screen

Jul 2026
Pinterest MLE coding round, one problem that looks like a classic LC but with a twist that changes everything. Took me a bit too long to realize the left-to-right evaluation rule was the whole point of the problem.
  • Given a string of digits and a target integer, insert +, -, or * operators between digits (or no operator to form multi-digit numbers) so the resulting expression evaluates to the target. Operators have no precedence and the expression is evaluated strictly left-to-right. Return all valid expressions. Multi-digit numbers cannot have leading zeros.

“I'd seen the LC 282 version before so I jumped straight into a backtracking approach and felt pretty good.”

View Post

Pinterest·Machine Learning Engineer·Technical Phone Screen

Jul 2026
Pinterest ML Engineer interview with a coding round that leaned more data-structures-heavy than I expected. The main problem was about tracking people moving through numbered rooms and answering count and ranking queries efficiently. Solid problem, took me a while to get the full solution clean.
  • Design a data structure to track n people moving through n rooms (all starting in room 0). Support two operations: count(roomId) returning the number of people in that room in O(1), and topKFastest(k) returning the k people who have moved the farthest, with ties broken by who arrived in their current room earliest.

“The count part came to me fast, just a counter array per room.”

View Post

Pinterest·Machine Learning Engineer·Recruiter / HR Screen

Jul 2026
Interviewed for an ML engineer role at Pinterest, and from what I can tell it was pretty standard screening territory.
  • Why do you want to work at Pinterest?

“Classic opener.”

View Post

Pinterest·Product Manager·Onsite - Product Sense / Strategy

Jul 2026
Pinterest product sense interview, one question about designing a shopping experience but for Netflix instead of Pinterest which felt like a weird pivot. Not a lot of context given upfront.
  • How would you design a shopping experience for Netflix?

“My first instinct was to ask why Netflix specifically, since Pinterest is literally a shopping-adjacent product and this felt like a curveball.”

View Post

Pinterest·Machine Learning Engineer·Technical Phone Screen

Jul 2026
Pinterest MLE interview with a string manipulation problem that sounds deceptively simple until you realize floats will absolutely betray you. The constraint about not using float() or Decimal was the whole point.
  • Given two numeric strings, one representing a value and one representing a precision, return the value rounded to that precision as a string. For example, rounding '12567' to the nearest '100' gives '12600', or rounding '1234.678' to '0.1' gives '1234.7'. You cannot use float() or Decimal helpers; implement the rounding directly on the string representations.

“My first instinct was to just cast to float and call it a day, and the interviewer shut that down immediately.”

View Post

Pinterest·Product Manager·Onsite - Product Sense / Strategy

Jul 2026
PM interview at Pinterest with a product strategy question about whether Slack should expand into the education market. Single question, no fluff, just had to make a case.
  • Should Slack enter the education market by launching a product specifically for education?

“I went straight into market sizing and user segmentation (students vs teachers vs admins) and it felt like the right move until I realized I hadn't actually argued for or against the opportunity clearly.”

View Post

Pinterest·Machine Learning Engineer·Technical Phone Screen

Jul 2026
Pinterest ML Engineer interview with a coding round that leaned more algorithmic than I expected. One problem stood out and I'm still thinking about whether my approach was actually clean or just lucky.
  • You're given an integer array called target. Starting from an all-zeros array of the same length, each operation lets you pick any subarray and increment every element in it by 1. What's the minimum number of operations needed to reach the target array?

“The key is realizing each operation maps to a contiguous range of increments, so you're really counting how many times the value goes up from one index to the next.”

View Post

Pinterest·Software Engineer·Technical Phone Screen

Jul 2026
Pinterest SWE interview with a string manipulation problem that looks dead simple but has a few layers once they start asking follow-ups. The core coding part was fine, the discussion afterward is where it got interesting.
  • Given two strings, determine whether the target string can be constructed using characters from the source string, where each character in the source can only be used as many times as it appears.

“I went straight to the frequency map approach, decrement counts as you consume characters, return false if anything dips below zero.”

View Post