Amazon Interview Questions & Trends 2026 Q2: OA, VO, Coding & Behavioral
Overall Trend
Amazon remains one of the most-discussed companies on the interview board. Recent threads show a stable interview structure but a clearly evolving OA format: alongside the traditional 2-question coding section, a new AI assistant bug-fix coding question has been added, plus Work Simulation (email scenario) and Work Style Surveys (personality) modules. VO is still BQ + Coding on every round, with BQ taking up a large share. GenAI usage BQ shows up in nearly every round. Intern flow: OA → 2 back-to-back VO rounds; some NG/SDE2 candidates get an extra HM evaluation call; L5+ Applied Scientist runs a 5-round onsite (Behavioral / Coding / HM / ML breadth / ML depth). [hidden]
Question Type Distribution
| Type | Count | Share |
|---|---|---|
| General Coding | 42 | 60% |
| ML Coding / Concepts / Debugging | 11 | 16% |
| System Design | 6 | 9% |
| Behavioral / Other | 11 | 15% |
Top 10 Hot Topics
| Rank | Topic | Count | Sample threads |
|---|---|---|---|
| 1 | OOD / scenario-style coding | 7 | thread1, thread2, thread3 |
| 2 | LRU Cache (LC 146) | 4 | thread1, thread2, thread3 |
| 3 | AI Assistant Debug (new OA format) | 4 | thread1, thread2, thread3 |
| 4 | [hidden] | [hidden] | [hidden] |
| 5 | [hidden] | [hidden] | [hidden] |
| 6 | [hidden] | [hidden] | [hidden] |
| 7 | [hidden] | [hidden] | [hidden] |
| 8 | [hidden] | [hidden] | [hidden] |
| 9 | [hidden] | [hidden] | [hidden] |
| 10 | [hidden] | [hidden] | [hidden] |
Question Index
General Coding
Log in to continue reading the full content
ML Coding / Concepts
Log in to continue reading the full content
System Design
Log in to continue reading the full content
Difficulty Trend
Algorithm bar has not visibly risen — most coding stays in easy-to-medium territory — but the format is shifting from pure leetcode to scenario-style OOD and AI-assisted engineering.
- Coding trends toward OOD / scenario problems. [hidden]; input format, edge cases, and tradeoffs are explicit scoring axes.
- OA adds an AI-debug task (60-min repo + AI assistant). [hidden]
- Phone Screen occasionally goes off-script: Applied Scientist saw a bucket-batching DP ([hidden]); Audible served a streaming abusive-books detection problem.
- ML interviews test breadth + depth. [hidden]; reciting concepts is far from enough.
Log in to continue reading the full content
Notable Changes
- New AI-Debug OA question type: [hidden]
- GenAI BQ is nearly mandatory: [hidden]
- HM Evaluation Call as a process step: [hidden]
- Recruiter ghosting is common: [hidden]
- 12-month cool-down: [hidden]
- More OOD / scenario problems: [hidden]
Log in to continue reading the full content
