Interview Questions Asked at Top Tech Companies
Last updated: June 2026
Every question in the LeakCode database was reported by a candidate who completed an actual interview at a specific company. These are not practice problems or hypothetical questions. They are real prompts reported by real candidates from their interview experiences. Select a company below to browse questions actually asked in its interview process.
How Question Attribution Works
When a candidate posts on 1Point3Acres, Blind, or Glassdoor about their interview experience, they typically name the company, the role, the round, and the specific questions they were asked. LeakCode parses these reports and attributes each question to the company and round where it was asked.
Attribution confidence varies. A question confirmed across multiple independent reports at the same company has high attribution confidence. A question appearing in a single post with minimal context has lower confidence. LeakCode surfaces cross-source frequency as a proxy for confidence: questions asked consistently are more reliably attributed.
Some questions appear at multiple companies. A graph traversal problem might be asked at Google and Amazon independently. These are attributed to each company separately. The pattern signal is still valid: both companies value graph proficiency even if the specific instances are different problems.
What Gets Asked in Tech Interviews
Candidate reports from the 25 companies on this page reveal consistent patterns in what gets asked across the industry:
-
Coding questions dominate the database for all major tech companies. Arrays, trees, graphs, dynamic programming, and string manipulation are the most reported categories. Difficulty skews medium to hard across all 25 companies.
-
System design rounds are standard at mid-level and senior positions. Candidates report being asked to design distributed systems, APIs, databases, and product features. The specific systems mirror each company's product surface.
-
Behavioral rounds use structured formats at most major companies. Amazon's leadership principles, Google's Googleyness criteria, and Meta's values all generate specific behavioral question clusters visible in the database.
-
Online assessments are the entry gate at most of the 25 companies. OA questions are time-sensitive data because prompts rotate frequently. The database is updated daily to reflect the most recent OA cycles.
Browse by Company
Select a company to see verified questions asked in its interview process, sortable by recency, role, and round type.
How Interview Formats Differ Across Companies
Amazon runs a longer loop than most, typically 5-7 rounds with heavy behavioral weight tied to leadership principles. Google uses a calibration committee model with multiple independent coding and system design rounds. Meta focuses on product sense and move-fast culture signals. Microsoft varies significantly by team but consistently runs behavioral and coding rounds.
Stripe and Bloomberg are known for especially rigorous technical screening. Stripe emphasizes code quality and pragmatic engineering judgment. Bloomberg has a distinctive multi-day in-person process. Doordash and Uber weight product-sense questions more than pure algorithm rounds, especially for senior engineering roles that touch product decisions.
Frequently Asked Questions
How do I know these questions were actually asked at a specific company?
Each question in the database comes from candidate reports that name the company, role, and round. Cross-source confirmation strengthens attribution confidence: when the same question appears in multiple independent reports at the same company over time, the attribution is high confidence. LeakCode surfaces frequency data so you can apply your own confidence threshold.
Do companies ask the same questions to every candidate?
No, but question pools are finite and recycled. A company might have 50-200 active coding questions rotating across interviewers at any given time. Candidates who interview in the same quarter frequently encounter overlapping problems. This is why recent high-frequency questions are the highest-value preparation targets.
Are the questions in the database too old to be useful?
Older questions have pattern value even if the exact prompt has rotated out. A company that asked a specific graph problem in 2022 and 2023 likely still values graph proficiency even if the exact problem changed. Use recent questions for specific prep and older questions for pattern calibration. The year filter in LeakCode lets you control the recency window precisely.
What is the difference between questions asked at a company and practice problems?
Practice problems (LeetCode, HackerRank, etc.) are designed to teach concepts. Questions asked at specific companies come from real interviews and reflect actual interviewer preferences, difficulty calibration, and follow-up patterns. Practice problems build foundational skills; company-specific reports calibrate you to what a specific employer actually tests in practice.