Amazon's SDE intern process is efficient: after the OA you go into two Virtual Onsite rounds, with no further screening. The timeline runs about a month—apply in late October, receive the OA in early November (two weeks to complete), and get the VO invite in late November. The interview runs over Amazon's own video tool Chime, both rounds are a behavioral + coding combo, and the BQ is 100% built on the 16 Leadership Principles. The two rounds together largely decide the outcome, so both carry high weight.
Amazon SDE Intern Interview at a Glance
| Dimension | Details |
|---|---|
| Process | OA → two VO rounds |
| Tool | Amazon Chime (download and test in advance) |
| Per-round duration | ~45 minutes |
| Per-round structure | Intro + BQ + coding + QA |
| Key | BQ is all built on the 16 Leadership Principles |
Each round splits into behavioral + technical: small talk and a self-intro, then BQ, then coding. The two rounds are similar in style but differ in problems and emphasis.
Round 1: Two LP Questions + Merge Intervals
1.1 Behavioral (~20 min)
Both BQs orbit the Leadership Principles, structured with STAR.
Question 1: "Tell me about a time when you had a conflict with a teammate and how you resolved it." Tests interpersonal handling and conflict management. I described overlapping responsibilities from unclear division of work: I proactively clarified task boundaries, guided the team to re-scope the plan, and moved the work forward while improving collaboration.
Question 2: "Tell me about a time when you had to learn something very quickly to complete a task." Tests rapid learning under pressure. I shared being asked to build with an unfamiliar framework on short notice: I read the docs and open-source projects fast and delivered the key module in two days.
1.2 Coding: Merge Intervals (Verbal Problem, No Run Environment)
Amazon's coding is distinctive: no on-screen problem; the interviewer dictates the requirements, and you must understand it, clarify ambiguities, explain your approach, write code, and describe testing with no run environment. The editor is bare—no test cases, no run button.
Problem: given a set of possibly overlapping intervals, merge all overlapping ones and return the result.
Reframe: sort + sweep line. Sort by start, iterate, merge when the current interval overlaps the previous, otherwise append.
def merge_intervals(intervals):
if not intervals:
return []
# Must sort by start first, or a single sweep can't merge correctly
intervals.sort(key=lambda x: x[0])
merged = [list(intervals[0])]
for s, e in intervals[1:]:
if s <= merged[-1][1]: # overlaps the previous interval
merged[-1][1] = max(merged[-1][1], e)
else:
merged.append([s, e])
return merged
Verbal points: explain why sorting is required (it makes overlapping intervals adjacent) and the merge boundaries (do touching endpoints count, empty input). Simulate the code mentally and describe outputs with your own test cases. Time: O(n log n). Space: O(n).
Round 2: LP Question + Number of Islands
2.1 Behavioral (~15-20 min)
Round 2's BQ leaned Ownership / Bias for Action:
Question: "Tell me about a time when you took ownership of a problem that wasn't strictly your responsibility." I described spotting an overlooked monitoring blind spot during my internship, proactively adding alerts and documentation, and driving the team to adopt it—emphasizing "take initiative + drive it to landing."
2.2 Coding: Number of Islands (Also Verbal, No Run Environment)
Problem: given a 2D grid of '1' (land) and '0' (water), count the number of islands. An island is surrounded by water and formed by connecting adjacent land horizontally or vertically.
Reframe: classic grid BFS / DFS connected components. Scan each cell; on unvisited land, increment the count and BFS/DFS-mark the whole island as visited.
from collections import deque
def num_islands(grid):
if not grid or not grid[0]:
return 0
rows, cols = len(grid), len(grid[0])
count = 0
for r in range(rows):
for c in range(cols):
if grid[r][c] == '1': # found a new island
count += 1
grid[r][c] = '0' # mark visited to avoid double counting
q = deque([(r, c)])
while q:
x, y = q.popleft()
for dx, dy in ((1, 0), (-1, 0), (0, 1), (0, -1)):
nx, ny = x + dx, y + dy
if 0 <= nx < rows and 0 <= ny < cols and grid[nx][ny] == '1':
grid[nx][ny] = '0'
q.append((nx, ny))
return count
Verbal points: explain why you mark an entire landmass as soon as you find one island (otherwise you double count), boundary checks (out of bounds, empty grid), and whether modifying the input grid is allowed (if not, use a separate visited set). Time: O(rows × cols). Space: O(rows × cols) (worst-case queue size).
Lessons Learned
| Lesson | Detail |
|---|---|
| BQ is all LP | Both rounds have BQ—prepare STAR stories across multiple dimensions |
| Articulation beats code | No run environment, so explain logic, boundaries, and test cases verbally |
| Time management | Each 45-min round fits intro + BQ + coding + QA—don't over-detail the BQ |
Rehearse the full flow's pacing rather than each piece in isolation—over-detailing the BQ squeezes your coding time.
FAQ
How many rounds is the Amazon SDE intern VO?
Two Virtual Onsite rounds. After the OA you get two rounds, each ~45 minutes and each a BQ + coding combo over Chime. Together they largely decide the outcome, so both BQ and coding must perform consistently.
Why is there no run environment in Amazon's coding round?
The interviewer dictates the problem; the editor is minimal with no execution and no test cases. This forces correct, clear code plus the ability to simulate it mentally and describe outputs and edge cases out loud.
What do the two BQ rounds ask, and how do I prep?
All built on the 16 Leadership Principles—common ones include conflict resolution, rapid learning, ownership, and bias for action. Prepare STAR stories spanning Ownership, Customer Obsession, Bias for Action, and more.
How do I manage time in each 45-minute round?
Keep the intro short, time-box the STAR-structured BQ, and leave ample coding time. Over-detailing the BQ crowds out coding—rehearsing the full flow's pacing helps most. If useful, our VO support / interview assistance can run two-round timed mocks at the real pace.
Preparing for the Amazon SDE intern VO?
The Amazon internship is two VO rounds, where BQ (LP) matters as much as verbal coding. If you want timed mocks of verbal-coding problems like merge intervals / number of islands, LP story polishing, or a full two-round pacing debrief, reach out: share the job description so we can predict the problems and BQ, then plan practice, with live VO support / VO proxy / interview assistance pairing available.
Add WeChat Coding0201 to get Amazon intern VO problems and mocks.
Contact
- WeChat: Coding0201
- Email: [email protected]
- Telegram: @OAVOProxy