Just finished it recently Amazon For the latest batch of OA on April 13, the standard configuration of Amazon OA is still 2 coding questions + 70 minutes (work simulation will be added for some roles). The platforms are mostly HackerRank or internal systems. The following is a summary of the two questions we did this time, which are also a very classic set of questions, with core ideas and speed-passing methods.
题目 1：Bug 报告优先级排序
Topic
你需要帮助亚马逊质量保障工程师处理自动化测试生成的 bug 报告日志。每条日志包含一个整数 bug 代码，同一次测试会话中可能出现重复的 bug 代码（同一问题被多次触发）。
为了高效排优先级，遵循以下规则：
- 出现频率越低的 bug 优先级越高（代表罕见 / 边界问题）
- 若两个 bug 出现次数相同，代码数值更小的 bug 优先级更高
任务：按照优先级从高到低对 bug 代码排序。
Example
Enter:bugs = [8, 4, 6, 5, 4, 8]
统计频率：
Sheet
| Item Code | Frequency |
|---|---|
| 8 | 2 |
| 4 | 2 |
| 6 | 1 |
| 5 | 1 |
Sorting rules:
- 频率 1 的 bug 优先于频率 2 的 bug
- 同频率下，代码小的优先：
5 < 6,4 < 8最终输出：[5, 6, 4, 4, 8, 8]
Constraints
1 ≤ n ≤ 2 * 10^51 ≤ bugs[i] ≤ 10^6
Problem-solving ideas
- Statistical frequency: Use a hash table (dictionary) to count the number of occurrences of each bug code
- Custom sorting: The sorting key is
(frequency, code), sorted by ascending frequency, ascending code - Restore array: In the sorted order, repeat each bug code according to its frequency to generate the final array
Topic 2: High-frequency product combination
Title
亚马逊零售分析团队需要找出最常被一起购买的商品对，用于生成 “组合购买” 捆绑包。
在观测窗口内，每个客户订单记录为空格分隔的 SKU 字符串（如"B07 B08 B09").
- 同一订单内，相同 SKU 重复出现仅计 1 次
- 任务：找出在最多订单中同时出现的不同 SKU 对；若多个组合并列第一，返回字典序最小的对（先比较第一个 SKU，再比较第二个）
- 最终返回按升序排列的两个 SKU
Example
Enter:orders = ["B07 B08 B09", "B07 B08", "B08 B09"]
订单内组合：
- 订单 1：
{B07,B08}, {B07,B09}, {B08,B09} - 订单 2：
{B07,B08} - 订单 3：
{B08,B09}全局计数： {B07,B08} → 2{B07,B09} → 1{B08,B09} → 2并列最高频，字典序B07 B08更小，最终输出["B07", "B08"]
Constraints
1 ≤ n ≤ 10^5- 每个 SKU 为 1-10 个字母数字字符
1 ≤ len(orders[i]) ≤ 50- 每个订单至少 2 个、最多 15 个不同 SKU
Problem-solving ideas
- Remove duplicate order SKU: For each order, first convert the SKU into a set to remove duplicates, and then sort (guaranteed combination
(a,b)MiddleA < b, to avoid duplication of statistics(b,a)) - Generate combined pairs: For each order’s SKU list, generate all
C(k,2)Combinations (KIt is the number of different SKUs in the order, the maximum is 15, and the maximum number of combinations is15*14/2=105, the total complexity is controllable) - Statistical combination frequency: Use a hash table to count the number of occurrences of each combination
- Screen the best combination: Traverse all combinations and find the pair with the highest frequency and smallest lexicographic order
Amazon OA Effective Preparation Recommendations
- Focus on LeetCode Amazon high frequency (Medium-Hard in the last 6-12 months), giving priority to Greedy, Prefix, and Heap.
- Do 2 questions strictly for 70 minutes, easy first and then difficult, leaving 10 minutes for boundary checking and optimization.
- Common pitfalls: edge cases (empty array, -1 cannot be satisfied), strict output format, floating point precision.
If you are preparing for Amazon SDE1, SDE2 or Intern 2026 OA, after you have mastered these high-frequency questions, your passing rate will be significantly improved. The overall thinking of the April questions is clear, but requires strong temporary analysis skills.
If you need detailed ideas for these questions, complete Python/Java code, more variations, or if you are pressed for time and want to pass the level efficiently, please feel free to contact us Programhelp OA ancillary services. We provide OA ghostwriting, VO real-time idea guidance and full-process package solutions, and have helped a large number of students successfully pass OA from major companies such as Amazon.
