← Back to blog
OA Amazon

Amazon NG OA 2026 Latest | SDE New Grad Online Assessment Full Walkthrough

2026-05-10

One-line summary: Amazon NG OA 2026 = Work Simulation (70 min) + Coding (70 min, 2 questions) + Work Style Survey (~10 min). The hardest piece is the second coding question — you need an O(n log n) optimal solution.

Amazon is one of the largest North-American SDE NG hirers. Invitations roll out fast, the question pool is fairly stable, and the OA is well worth grinding because "if you've practiced, you'll pass." This article documents the latest 2026 (April-May) reports.

1. Amazon NG OA Module Layout

Module Duration Questions Notes
Work Simulation ~70 min Multiple scenarios Real-work judgment
Coding ~70 min 2 Decisive
Work Style Survey ~10 min 30+ Consistency check

Important: The three modules must be completed back-to-back (one allowed pause). Results come back in 7-15 days.

2. Coding — 2 Latest Real Questions

Q1: Server Maintenance Window

Statement: Given each server's busy intervals, find the longest contiguous window when all servers are simultaneously idle.

Idea: sweep line + interval merge

def maxMaintenanceWindow(intervals_list, T):
    events = []
    for intervals in intervals_list:
        for s, e in intervals:
            events.append((s, 1))
            events.append((e, -1))
    events.sort()

    busy = 0
    last_free_start = 0
    best = 0
    for t, delta in events:
        if busy == 0:
            best = max(best, t - last_free_start)
        busy += delta
        if busy == 0:
            last_free_start = t
    best = max(best, T - last_free_start)
    return best

Complexity: O(n log n)

Q2: Inventory Optimization (the famous "13-min AC")

Statement: A warehouse has n slots with volumes v[i] and prices p[i]. Knapsack capacity is C, but adjacent slots may be merged into a new slot whose volume is the sum and value is the maximum of the two. Find max total value.

Idea:

def maxValue(v, p, C):
    n = len(v)
    items = list(zip(v, p))
    for i in range(n - 1):
        items.append((v[i] + v[i+1], max(p[i], p[i+1])))

    dp = [0] * (C + 1)
    for vi, pi in items:
        for c in range(C, vi - 1, -1):
            if dp[c - vi] + pi > dp[c]:
                dp[c] = dp[c - vi] + pi
    return dp[C]

Pitfalls:

3. Work Simulation Module

Sub-module Time Content
Day in the Life 25 min Multi-decision day
Project Management 20 min Prioritize tasks
Code Review 15 min Pick LP-aligned PR
Customer Email 10 min Write/select reply

Scoring: Amazon Leadership Principles (LPs) drive consistency scoring.

Tip: Always favor Customer Obsession + Bias for Action + Ownership. Avoid "wait for manager" or "delay to next standup" options.

4. Work Style Survey

Purpose: Detect consistency. Each LP appears 3-4 times in different wordings. Inconsistent answers get flagged "untrustworthy." Pre-write 2-3 LP self-descriptions.

5. Amazon NG OA Prep Path (3 Weeks)

Week Focus Daily output
1 LeetCode Amazon Tag Top 50 5-7
2 DP + interval sweep + binary search 5 + 1 real Q
3 LP stories + Work Simulation mocks 8-10 STAR stories

6. FAQ — Amazon NG OA Common Questions

Q1: How long does the OA take to come back?

Usually 7-15 business days. If nothing by day 14, you can politely follow up with the recruiter.

Q2: Do I need to AC both coding problems?

Aim for both ACs. AC'ing only one still has roughly a 20% VO chance, but missing both is essentially auto-rejection.

Q3: How important are Work Simulation & Survey?

Very. Coding-perfect candidates with poor LP alignment do get rejected. Amazon really does score all three modules holistically.

Q4: Can I retake the OA?

No. Once per year per role. Reapplying inherits the previous score.

Q5: Python or C++?

Python is plenty. Time limits are friendly to Python (n ≤ 10⁵), and writing speed beats execution speed here.

Q6: What's the post-OA process?

OA pass → 1 Phone Screen (45-min coding + 15-min BQ) → 4-round Loop (2 Coding + 1 SD/Code Review + 1 Bar Raiser).

7. Top 8 LPs to Memorize

LP Keywords
Customer Obsession User-first, long-term value
Ownership No blame, long-term mindset
Invent and Simplify Open to outside ideas
Are Right, A Lot Data-driven
Learn and Be Curious Continuous learning
Insist on the Highest Standards No quality compromise
Bias for Action Speed
Deliver Results Quantified outcomes

8. External Resources

🚀 Need Amazon NG OA / VO Coaching?

If you're preparing for Amazon NG SDE Loop (with Bar Raiser), we can help with LP story polishing, coding mocks, and SD direction.

👉 Add WeChat: Coding0201 — get real questions and a 1-on-1 prep plan

Contact