技术面试英语 — Tech Interview English

读完本页后,你应该能够:

• 用 STAR 法(Situation/Task/Action/Result)讲行为面试题,2-3 分钟讲清楚
• 在 coding 题里用 think-aloud chunks 让 interviewer 跟得上你的思路
• 用 clarifying questions 的固定句式把题面信息确认完整
• 在 system design 题里用 scale → bottleneck → trade-off 的顺序展开
• 用 graceful chunks 处理没听懂、不知道、需要时间思考三种尴尬
• 谈薪时不直接给数字,用 "I'm looking for a range" 等技巧争取最高
• 识别并避免中国候选人 5 种常见 Chinglish 模式

每阶段考的不一样,英语模式也不同。下面 §2-§6 各拆一节。

2.1 STAR 框架

每个行为题答 2-3 分钟,严格按 4 段:

2.2 高频 STAR opener chunks

Situation:
  "Back when I was working on [project] at [company], we were facing [challenge]..."
  "About a year ago, my team was responsible for [thing] and we hit [problem]..."
  "Let me give you an example from my time at [previous role]..."

Task:
  "My specific role was to [task]..."
  "I was the technical lead on [aspect]..."
  "I was tasked with [specific responsibility]..."

Action:
  "The first thing I did was [step 1]. Then I [step 2]..."
  "My approach was three-fold: [a], [b], and [c]..."
  "I started by [investigating / proposing / building]..."

Result:
  "As a result, we [outcome] — that translated to [metric]."
  "We ended up [outcome], which [impact on business / team]."
  "The change cut [thing] by X%, which [why it mattered]."

2.3 高频问题套路应对

Coding 不是"在白板上写出对的代码",是"展示你的思考过程"。沉默 = 失败,即使代码对了 interviewer 也打低分。

3.1 Clarifying questions(题面 → 边界)

听完题先问 3-5 个 clarifying question。这一步和写代码同样重要:

Input / Output:
  "Just to make sure I understand — the input is X, output is Y, right?"
  "Are there any constraints on the input size?"
  "Can the input be empty / null?"
  "What's the expected behavior if [edge case]?"

Assumptions:
  "I'll assume [thing] for now — let me know if that's wrong."
  "Should I assume the input is sorted / unique / non-negative?"

Output format:
  "Should the function return X or print Y?"
  "Is in-place modification allowed, or should I return a new structure?"

3.2 Think-aloud while coding

写代码时每一步说出来:

Approach:
  "My initial thought is to use [data structure / algorithm]. Let me explain why..."
  "There are two ways to approach this: [A] or [B]. A is simpler but B is faster."
  "I'll start with the brute force, then optimize."

Mid-solve:
  "Now I need to handle the edge case where [thing]..."
  "Wait, I think I see a bug — let me trace through..."
  "Hmm, this is O(n²). Let me think if there's a way to do it in O(n log n)..."

Stuck:
  "I'm a bit stuck on [thing]. Let me think out loud — [hypothesis]..."
  "Could I get a small hint? I'm not sure if [direction] is right."

3.3 没听懂 / 不知道时的 graceful chunks

System design 题答 45-60 min,考的是"工程判断"。结构最重要:

4.1 标准展开结构

1. Clarify (5 min)
   - "What's the scale? DAU? requests / sec?"
   - "What features must we support? What's out of scope?"
   - "Read-heavy or write-heavy? Latency requirements?"

2. High-level design (10 min)
   - 画 component 图: client → CDN → LB → app servers → cache → DB
   - 用 chunks: "Let's start with the core flow: when a user [does X], the request goes to [component]..."

3. Deep dive on 2-3 components (20-30 min)
   - DB schema + indexing
   - Cache strategy (write-through? write-back? TTL?)
   - Sharding / replication

4. Trade-offs (5-10 min)
   - "I chose [X] because of [reason], but it has [downside]. The alternative would be [Y]."

5. Wrap up (last 5 min)
   - "If we had more time, I'd also discuss [monitoring / failure modes / cost]."

4.2 高频 chunks

Quantifying scale:
  "Let me estimate the QPS first — assuming X DAU, Y actions per day per user..."
  "That gives us roughly Z reads per second and ~Z/10 writes per second."

Component reasoning:
  "We need a cache here because [reason]. I'd use Redis for [property]."
  "For the DB, I'd choose [SQL / NoSQL] because the access pattern is [reason]."

Trade-offs:
  "The trade-off here is consistency vs availability. Given the use case, I'd lean toward [choice]."
  "Strong consistency would simplify [thing] but cost us [thing]. Eventual is fine for this use case because [reason]."

Handling follow-up:
  "Good question — let me think about that. If [scenario], we could [approach]..."
  "That's a great point. To handle that, I'd add [component / mechanism]..."

Culture fit 考"你和公司价值观是否对得上"。常见为 Amazon LP / Google googliness / Meta move fast。

5.1 答题原则

• 具体例子(STAR 法适用),不是空 talk
• 诚恳承认短板,不要伪装完美
• 提问反向: "Could you tell me more about how the team handles [X]?" 让对方知道你认真在评估

5.2 高频 chunks

Why this company:
  "I've been following [company] since [time]. What stands out is [specific thing]."
  "The reason I applied is [specific reason] — not just the brand."

Cultural alignment:
  "I value [specific principle], and I think it aligns with your [specific value/practice]."
  "From what I've read about [team / OKR], it seems like a great fit because..."

Asking back:
  "What does success look like in this role at the 6-month mark?"
  "Could you walk me through how the team makes technical decisions?"
  "What's the team's biggest current challenge?"

谈薪是面试最被中国候选人低估的环节。90% 的人在第一轮就说出数字 → 失去主动权。原则:让对方先报。

6.1 引导对方先报价的 chunks

When asked "What's your expected salary?":
  "I'd love to know more about the full comp package — base, bonus, equity — before sharing a number."
  "I'm flexible on numbers but I'm focused on finding a role with strong [tech / growth / ownership]. What does the band for this level look like?"
  "I'm currently at [range], but I'm looking for a step up given [my growth / market]. What's the budget for this role?"

If pushed harder:
  "Based on my research and current offers, I'm looking in the range of $X to $Y, but I'm open to discussing the full package."
  (Give a range; the bottom of YOUR range should be ~10-15% above what you'd actually accept.)

6.2 收到 offer 后的 negotiation chunks

Anchoring:
  "I'm really excited about the offer. Before I commit, I want to discuss the comp."
  "I have another offer at $Z — I'd love to make this work but the gap is significant."

Specific asks:
  "Would you be open to increasing the base by $X? That would put it more in line with [market data / other offer]."
  "Could we look at the equity component? I'd be more comfortable at [N] shares given the 4-year vest."
  "Is there any flexibility on the sign-on bonus?"

Closing:
  "If you can do [specific number], I'd sign today."
  "Let me think about it overnight and get back to you tomorrow morning." (Always OK to take time.)

6.3 避免的 5 个错误

• 行为面试用 STAR 法:Situation/Task/Action/Result 4 段,2-3 min 讲完
• Coding 必须 think-aloud —— 沉默 = 失败,即使代码对也低分
• System design 按 Clarify → High-level → Deep dive → Trade-offs 4 步展开
• Clarifying questions 是题面理解的关键,3-5 个 question 把边界问清
• Culture fit 要给具体例子 + 真诚提问反向,不是空 talk
• 谈薪让对方先报价,自己永远说 range 不是单一数字,offer 100% negotiate
• 5 种 Chinglish 模式:"experience on" / "responsible to" / "purpose is to" / "try my best" / "maybe I can"——逐个修
• 面试前 1 周 prep 计划:STAR 故事 / coding / mock / 公司研究 / 重过 / 放松

• ← 索引
• Meeting & Presentation chunks
• Business Email
• Engineering Writing
• Academic Vocabulary
• Pronunciation 中国 accent 弱点
• IELTS Speaking
• Advanced Collocations