← Back to Directory

TikTok interview notes, page 8 of 28

Large Enterprises

Most of these TikTok write-ups come from Data Scientists (111), then Software Engineers (102) and Product Managers (35). The most reported rounds are the technical phone screen (167), the product sense round (33), the system design round (25) and the behavioral / leadership round (21).

279 interview notes · updated Sep 2026

TikTok·Data Scientist·Technical Phone Screen

May 2026
Brutal TikTok DS interview question covering basically every dimension of experimentation at once. The kind of prompt where you finish answering and still aren't sure if you got even half of it right.
  • You're launching a tipping feature on creator posts and need to design an A/B test that handles interference and supply-demand dynamics. Walk through your randomization unit and clustering strategy to control for cross-user and cross-post spillovers.
  • How would you define eligibility and exposure rules for this experiment to prevent treatment contamination across different time zones, specifically US and Asia?
  • Define the primary metric hierarchy for this tipping experiment, including guardrail metrics around retention, session length, abuse, and ad revenue cannibalization.
  • How would you approach power analysis and experiment duration for this test, including ramp planning, sample ratio mismatch detection, and pre-registered stop rules?
  • What variance reduction techniques would you apply here, and how would you handle inference given clustered randomization?
  • If the tipping feature improves monetization metrics but causes a small but real drop in growth metrics, what decision framework would you use to decide whether to ship it?

“I went with creator-level randomization pretty quickly, reasoning that if a viewer sees a tipped post and then a non-tipped post from a different creator, you've already got contamination.”

View Post

TikTok·Data Scientist·Technical Phone Screen

May 2026
TikTok data scientist interview with a meaty offline marketing measurement question. The whole thing revolved around a single scenario but it had enough layers to keep you busy for a while.
  • Pinterest is running a billboard campaign in train stations to drive user traffic and engagement. How would you measure whether the campaign is actually working? Walk through how you'd define and quantify engagement tied to the billboards, estimate who's being exposed to them, and build a framework to forecast incremental gains like sign-ups, sessions, or revenue in the targeted areas.

“This was one question but it was basically four questions stacked on top of each other.”

View Post

TikTok·Software Engineer·Technical Phone Screen

May 2026
TikTok software engineering interview with a tricky string segmentation problem that had two parts, the second of which had a circular dependency baked into it that made things genuinely hard to reason about under pressure.
  • Given a message string and a width W, split the message into consecutive chunks of at most W characters each, then append a positional suffix like '1/3', '2/3', etc. to each chunk. The suffix does not count toward the width limit. Return the list of labeled segments.
  • Now modify the segmentation so the suffix IS included in the width limit. Since the number of segments affects the suffix length, and the suffix length affects how many segments you need, how do you resolve this circular dependency and find a valid segmentation efficiently? What's the complexity?

“Part A was fine, pretty mechanical.”

View Post

TikTok·Software Engineer·Technical Phone Screen

May 2026
TikTok software engineer interview with a string chunking problem that had two parts, the second of which was genuinely tricky. The easy part felt routine but part two had a circular dependency that took me a while to untangle.
  • Given a string and a width, split the string into chunks while preserving order, then append a suffix like '1/3', '2/3', etc. to each chunk. In part one, the suffix doesn't count toward the width. In part two, it does, and you don't know the total number of chunks upfront since the suffix length depends on it. Design an algorithm that handles both cases, covers edge cases, and explain complexity.

“Part one was fine, just divide the string into slices of size w and slap the suffix on.”

View Post

TikTok·Software Engineer·Onsite - Coding / Algorithms

May 2026
TikTok software engineer coding round with two problems, one pretty mechanical and one that requires you to actually think about data structures. Nothing too wild but the second problem has some gotchas if you're not careful.
  • Given two integer arrays A and B and a threshold T, for each element in A count how many elements in B satisfy a * b >= T. Return the counts as an array.
  • You have a list of user accounts where each account has a name and a set of identifiers. Accounts sharing at least one identifier belong to the same real user, including transitively. Merge them and return each merged group with its name and a sorted, deduplicated list of identifiers.

“Sort B and binary search for the cutoff point.”

View Post

TikTok·Data Scientist·Take-home Assignment

May 2026
TikTok data scientist interview with a practical Pandas/ecommerce scenario. The whole thing felt more like a take-home exercise than a live coding session, which I wasn't expecting.
  • Given a raw ecommerce sales table, use Pandas to filter for European orders placed in the last 30 days and compute their total revenue.
  • Identify and remove duplicate orders from the dataset, then describe how you would handle missing revenue values.
  • Generate summary statistics for the sales data and produce two visualizations that highlight trends across product categories.

“The filtering part was fine.”

View Post

TikTok·Machine Learning Engineer·Onsite - Multi Round

May 2026
TikTok MLE interview that's basically a deep-dive gauntlet on your resume projects. Four projects, full technical breakdown, and then they push back on your choices to see if you fold or fight.
  • Walk me through a significant project on your resume, covering the problem and business context, the constraints you worked under (latency, cost, privacy, fairness, etc.), your specific role and ownership, the architecture and key decisions, alternatives you considered, measurable outcomes, and the hardest challenge you faced.
  • For each project, what alternatives did you seriously evaluate, and how did you make the final call between them?
  • What were the quantified outcomes of your projects? Things like watch-time lift, latency improvements, cost reduction, revenue impact.
  • If the interviewer tells you your approach performs poorly or says they wouldn't do it that way, how do you respond? Can you defend your trade-offs with data, or would you revise the design?
  • Describe a time you received blunt or dismissive feedback during a design review or interview. What did you do in the moment and what changed afterward?
  • If an interviewer insists on discussing a different programming language or coding style than what you're comfortable with, how do you keep the conversation productive?

“This is the core of the whole interview and it took way longer than I expected.”

View Post

TikTok·Machine Learning Engineer·Technical Phone Screen

May 2026
TikTok ML Engineer interview with a dynamic programming problem. Pretty standard for this type of role but worth knowing cold before you go in.
  • Given an array of non-negative integers, find the maximum sum of a subset where no two selected elements are adjacent in the original array.

“Classic house robber style problem.”

View Post

TikTok·Data Scientist·Technical Phone Screen

May 2026
TikTok data scientist interview that was basically one very long case question about ad reporting. The whole thing was a deep causal inference problem and they wanted you to go end-to-end, from hypotheses to statistical models to quasi-experiments. Felt more like a take-home presented verbally.
  • You see a positive correlation between ad page visits and the rate at which ads get reported as bad. Walk through whether more visits actually cause more reports, or whether something else explains the relationship.
  • What specific metrics and data cuts would you use to diagnose whether the correlation is driven by exposure or by something else?
  • Build a statistical model to estimate the effect of page visits on report probability while controlling for exposure. What does your coefficient mean and how do you check for issues like overdispersion?
  • Describe a quasi-experimental approach you'd use to establish causality here, and how you'd validate it.
  • What thresholds would you set for triggering an operational alert, and what product or policy actions would you recommend depending on the outcome?

“This is a beast of a question.”

View Post

TikTok·Software Engineer·Technical Phone Screen

May 2026
TikTok software engineering interview with a pretty gnarly algorithmic problem that had multiple layers to it. The O(log n) constraint is what makes it interesting and also what makes it easy to freeze up on.
  • Given a sorted integer array of length n, find all distinct values that appear strictly more than n/3 times. Your solution must run in O(log n) time with O(1) extra space. As part of your answer: prove there can be at most two such values, explain how you'd use binary search to identify and verify candidates, argue correctness, analyze complexity, and handle edge cases like empty arrays or all-equal elements.

“The pigeonhole argument for 'at most two' came to me pretty fast, that part felt good.”

View Post