Name: 資料工程師面試必考題 Top 10+ 完整解答！ (Top 10+ Data Engineer Interview Questions and Answers)
Uploaded: 2026-03-28T09:27:40.000Z
Duration: 13 min 19 s
Description: 【看影片學英語】數萬部 YouTube 影片，搭配英漢字典即點即查，輕鬆掌握單字發音與用法，長久累積看電影不必再看字幕。

and today we're going to talk about the data engineering interview, example questions and solutions,

今天我們要來聊聊資料工程師的面試，像是常見問題跟解答，

情態助動詞 B1

and then go over the 8 different types of questions that might get asked for data engineers.

然後會帶大家看資料工程師面試中可能會遇到的 8 種問題類型。

We analyzed thousands of data engineering interview questions,

我們分析了數千題的資料工程師面試問題，

and we saw that there's pretty common types between the two.

然後發現這兩者之間其實有很多常見的類型。

Number 1: the most common interview question type is usually around behavioral interview questions,

第一種：最常見的面試問題類型通常是關於行為面試問題，

and behavioral interview questions get asked to basically every candidate.

行為面試問題基本上會問到每一位應徵者。

But for data engineers, they're going to be a little bit more specific towards data engineering and your previous projects.

不過對於資料工程師來說，這些問題會更貼近資料工程和您過去的專案。

For example, one question would be like, "Tell me about a time that you saw an inefficiency in data processing,

舉例來說，其中一個問題可能會是：「請分享一個你發現資料處理上有效率低下的情況，

and how did you solve it?" Another question would be like, "How would you work with a product manager to develop a new product?

然後你又是怎麼解決的呢？另一個問題可能會是：「你會怎麼跟產品經理合作開發新產品？

How would you consider the data pipeline side of it?" For these questions, they're really assessing just how you can work with others cross-functionally and other data engineers on your team.

在資料管線方面你會怎麼考慮？」這些問題是在評估你如何與他人跨職能地合作，以及與團隊中的其他資料工程師如何合作。

So a lot of this is just talking about your experience, really going into the details about your data engineering knowledge,

所以這大部分是在談論你的經驗，深入探討你的資料工程知識，

and making sure that you can communicate your projects and your experience well.

並確保你能清楚地溝通你的專案和經驗。

Let's look at a question in greater detail.

Let's say that the interviewer asks you to describe one of your data engineering projects that you worked on before.

假設面試官請你描述一個你之前做過的資料工程專案。

So with this question and a lot of other behavioral interview questions, the easiest way to use is the STAR format.

對於這個問題以及許多其他的行為面試問題，最簡單的使用方式就是 STAR 格式。

STAR stands for Situation, Task, Action, and Result.

STAR 代表情境、任務、行動和結果。

So describe the situation, what exactly happened with the data engineering project.

所以請描述情境，也就是這個資料工程專案到底發生了什麼事。

Were you basically creating an events database?

你當時是在建立事件資料庫嗎？

Then provide an overview of the actual task at hand.

What was the problem that you were trying to solve, and then how did you actually solve it?

你當時想解決什麼問題，然後又是怎麼解決的呢？

Then talk about the actions that you took.

What did you actually do as part of that project?

在這個專案裡，你實際做了些什麼？

Did you actually design the schema yourself, or did you work with another engineer to do so?

你是自己設計綱要，還是跟其他工程師一起合作完成的？

And then lastly, tell the interviewer about the results.

Did you improve the data pipeline and data accuracy by X percent?

你改善了資料管線和資料準確性多少百分比？

Did you improve the data quality and help increase the level of confidence in data to the executives?

你是否改善了資料品質，並幫助高階主管對資料的信心提升？

So stuff like that is really helpful when describing all your projects.

所以像這樣的內容，在描述你所有專案時都非常有幫助。

STAR is pretty basic, and you can read more about it below in the description.

STAR 格式很基礎，你可以在下方的說明裡閱讀更多相關資訊。

Number 2: a second pretty common question we see is just general basic technical questions around data engineering.

第二點：另一個很常見的問題，就是關於資料工程的基礎技術問題。

So these questions would be something like, "What ETL tools do you use?

所以這些問題可能會像是：「你使用哪些 ETL 工具？

What's your experience with cloud technologies, Azure, Spark, other things like that?" These questions are more so understanding your depth of experience using these tools because a lot of these companies might have specific workflows,

你在雲端技術、Azure、Spark 或其他類似方面的經驗如何？」這些問題主要是想了解你使用這些工具的經驗深度，因為很多公司可能有特定的工作流程，

and they like to prioritize candidates that have used these tools before.

他們喜歡優先考慮有使用過這些工具的應徵者。

So if you have experience in Airflow, then you do have an advantage.

所以如果你有 Airflow 的經驗，那你就具有優勢。

And then they're just kind of testing you really basic kind of requirements on that you knew how to use Airflow and that you actually used it before versus kind of putting it on your resume because you read about it once.

所以這些問題主要是想測試你是否真的知道如何使用 Airflow，而且你確實用過它，而不是因為你只看過一次就把它寫在履歷上。

The rule of thumb for these questions is just don't put anything on your resume that you actually haven't directly had any experience with yourself,

這些問題的經驗法則就是，不要把任何你沒有親身經驗過的事情寫在你的履歷上，

and then you could probably answer these questions pretty easily.

這樣你大概就能輕鬆回答這些問題了。

The third most common type of question is SQL,

第三種最常見的問題類型是 SQL，

and the SQL probably comes up 95 percent of the time.

而且 SQL 大概有 95% 的機率會出現。

It's probably the most common technical interview question that shows up for data engineers.

這大概是資料工程師最常遇到的技術面試問題了。

SQL questions are basically core to any data engineering job.

SQL 問題基本上是任何資料工程職位的核心。

You need to know the fundamentals for just joins, subqueries, case statements,

你需要知道關於 joins、subqueries、case statements 的基本功，

and you probably will be asked pretty basic level SQL questions up to really hard SQL questions.

而且你可能會被問到從非常基礎到非常難的 SQL 問題。

Most SQL questions exist to weed out candidates that just don't have any experience.

大多數 SQL 問題的用意是淘汰掉那些根本沒有經驗的應徵者。

And so if you've been a data engineer before, this should be pretty easy.

所以如果你之前是資料工程師，這應該會蠻簡單的。

If you haven't and you're a new grad looking to jump into data engineering, you can practice SQL questions on interview query and get really good at them.

如果你是剛畢業想進入資料工程領域的新鮮人，你可以在 interview query 上練習 SQL 問題，然後練到很熟。

Given a user's table, write a query to return only its duplicate rows.

給定一個使用者的表格，請寫一個查詢來找出重複的列。

One is named projects and another maps employees to the projects that they're working on.

一個叫做 projects，另一個是員工與他們正在進行的專案的對應關係。

Given the employees and departments table, write a query to get the top three highest employee salaries by department.

給定 employees 和 departments 表格，請寫一個查詢來找出每個部門薪水最高的前三名員工。

You're given a table that represents search results from searches on Facebook.

你得到一個代表 Facebook 搜尋結果的表格。

Position column represents each position the search result came in,

position 欄位代表搜尋結果出現的每個位置，

and the rating column represents the human rating of the result from 1 to 5, where 5 is high relevance and 1 is lower relevance.

rating 欄位代表人類對搜尋結果的評分，從 1 到 5 分，5 分表示高度相關，1 分表示較低相關。

Write a query to get the percentage of search queries where all the ratings from the query results are less than a rating of 3.

請寫一個查詢，找出搜尋查詢中所有搜尋結果的評分都低於 3 分的百分比。

Okay, let's dive into one of these questions and go over the solution as well.

好的，我們來深入探討其中一個問題，並且一起看看解答。

Given three tables representing customer transactions and customer attributes, write a query to get the average order value by gender.

給定三個代表客戶交易和客戶屬性的表格，請寫一個查詢來找出每個性別的平均訂單價值。

So we can see from these tables that we actually have to join all three of them likely because of the fact that we need to get the average order value, which is going to be a function of the product times the actual quantity times the price.

所以我們可以看到，從這些表格中，我們其實需要將它們全部連接起來，這很可能是因為我們需要取得平均訂單價值，而這將是產品乘以實際數量乘以價格的函數。

All we have to do is join users to transactions and then join that to products as well, and

我們只需要將 users 連接到 transactions，然後再將其連接到 products，

then group by the sex of the user and then get the average quantity multiplied by the price.

然後依據使用者的性別分組，接著取得平均數量乘以價格。

If you guys are looking for more resources for SQL interview questions, I've done a ton of videos of this in the past,

如果你們正在尋找更多關於 SQL 面試問題的資源，我過去已經拍了很多相關的影片，

and you can find them in the description below.

你們可以在下方的說明中找到它們。

The fourth and fifth most common interview questions get asked for data engineers are on data structures and algorithms.

第四和第五個最常問到資料工程師的履歷問題，是關於資料結構和演算法。

Now, these are actually bucketed into two groups.

現在，這些其實會抽成兩大類。

One is more so data structures and algorithms that you could expect on questions like LeetCode.

第一類比較像是你在 LeetCode 上會遇到的資料結構和演算法問題。

These are testing your ability with just being able to write coding and basically code as a software engineer.

這些是在測試你身為軟體工程師，能不能寫程式碼、實際寫出程式的能力。

The second type is more so common to actual data manipulation using Python or Pandas.

第二類則比較偏向使用 Python 或 Pandas 進行實際的資料處理。

And so this is basically taking data and manipulating in some way.

所以這基本上就是對資料進行某種方式的處理。

But it's not as algorithm focused as the LeetCode type problems are.

但它不像 LeetCode 那種類型的問題，那麼著重於演算法。

For example, you'd be doing something like using Python lists to iterate over date and time, basically munching different kinds of data types, converting ints to strings and strings back to ints,

舉例來說，你會做像是使用 Python 的 list 來迭代日期和時間，基本上是處理各種不同類型的資料，將整數轉換成字串，再將字串轉回整數，

and then doing some data munching in Pandas, basically doing some feature selection, maybe merging two datasets, stuff like that.

然後在 Pandas 中進行一些資料處理，像是特徵選擇，或是合併兩個資料集，諸如此類的事情。

So example of more of the data munching Python type questions are write a function called findByGrams to take a string and return a list of all the bygrams.

所以一個比較偏向資料處理的 Python 問題範例是，寫一個叫做 findByGrams 的函式，接收一個字串並回傳所有雙字詞的列表。

So here you're just doing a little bit of data munching.

所以這裡你只是做一點點資料處理。

You're basically going through a really long string and creating bygrams.

你基本上是遍歷一個很長的字串並建立雙字詞。

So basically kind of like feature selection that you'd expect in data science,

所以基本上就像你在資料科學中預期的特徵選擇，

but also shows up in data engineering as well.

The question that's probably more common to LeetCode would be something like given a list of integers and an integer n, write a function called sumToN that finds all the combinations that sum up the value to n.

一個比較常見的 LeetCode 問題可能是，給定一個整數列表和一個整數 n，寫一個叫做 sumToN 的函式，找出所有加總為 n 的組合。

So this is a pretty common LeetCode question.

所以這是一個相當常見的 LeetCode 問題。

Basically, you return a list of integers that sum up to this n target variable that you also input in.

基本上，你回傳一個整數列表，它們加總起來等於你輸入的目標變數 n。

Very common, data engineers are expected just to be able to code and code well.

非常普遍，資料工程師被期望能夠寫程式，而且寫得很好。

So this is going to be a basically common part of any interview that you go to.

所以這基本上會是你參加任何面試時都會遇到的常見部分。

And so this actually segues really nicely into our sponsor of today's video, which is Interview Query.

這也正好銜接到我們今天影片的贊助商 Interview Query。

So Interview Query is a startup that I founded that basically helps you with your interview prep for any kind of data engineering interview that you have coming up.

Interview Query 是我創辦的一家新創公司，主要幫助你準備即將到來的任何資料工程面試。

Primarily, we have tons of interview questions, courses, and

主要來說，我們有大量的面試問題、課程，以及

challenge assessments so you can see how you rank compared to your peers.

挑戰評估，讓你瞭解自己與同儕的排名。

Additionally, we also have an interactive code editor where you can run SQL and Python and basically view your test cases.

此外，我們還有一個互動式程式碼編輯器，你可以在裡面執行 SQL 和 Python，並查看你的測試案例。

And also, you could see solutions from other community members as well.

而且，你也可以看到其他社群成員的解決方案。

We're actually giving away one free year of Interview Query premium access to anyone who comments on this video talking about their interview experiences as a data engineer or just any questions they have about the data engineering interview.

我們將免費贈送一年的 Interview Query premium 權限給任何在這支影片下方留言，分享他們作為資料工程師的面試經驗，或是任何關於資料工程師面試的問題的人。