Hi all — long-time lurker, first-time poster.

I’m finally sharing my insights to help the community prepare for the data engineer interviews. I won’t dox myself revealing too much information, so please don’t ask for too many details.

Who am I? I’ve worked in the software/data space for nine years and I’m based in Europe, this summer I wanted to check the current state of data engineer interviews, see whether my skill set is lacking, and find out what the market is offering. Over the last few months I interviewed for several mid-size or lesser-known companies and for these faang:

1. Meta (London), Product Data Engineer, L5
2. Apple, Software & Data Engineer (team 1)
3. Apple, Software & Data Engineer (different team 2)
4. Apple, Analytics Data Engineer
5. Netflix (Warsaw), Senior Data Engineer, L5

I was ghosted by Amazon Prime Music (Data Engineer, London), Cloudflare (Data Engineer, remote — Europe), and Spotify (Senior Data Engineer, London)

Resume

My resume is a standard data-engineer CV — no FAANG pedigree or famous university — just nine years of experience in software and data. Having data-streaming expertise helped; it seems to be a differentiator right now. No special tips: write a clear, honest CV and list your accomplishments precisely.

Interview types

Data-engineer job titles and expectations vary widely by company and team.

• Meta distinguishes “Product” and “Infrastructure” data-engineer roles with different processes.
• Apple is highly team‑oriented: some teams focus on Spark, others on backend engineering or analytics.

Common interview areas:

• System design for data (e.g., design a click‑ingest pipeline)
• Data modeling (facts/dimensions, schema evolution, slowly changing dimensions)
• Python coding (I rarely saw heavy LeetCode-style questions; Meta was an exception)
• Behavioral interviews
• SQL (queries based on given schemas)
• Spark, distributed computing, and debugging
• Team-specific topics can include Java multithreading or other domain-specific skills.

Newer interview formats I encountered (possibly to deter AI cheating):

• Code comprehension: read a codebase (~10–15 Java files) and add a feature or fix a bug.
• Code refactoring: improve existing working code for maintainability or performance.

Meta

Phone screen
50 minutes: 3 Python and 3 SQL questions. Code must compile and pass tests; you can run it.
Questions are generally easy (LeetCode‑easy), but time pressure is significant.

Onsite
Rigid structure with short, focused interviews:
15 min product discussion (define metrics and what to track)
15 min ETL design (fact/dim, 1:many, 1:1 relationships)
15 min SQL (write queries from a given schema)
15 min Python (easy/medium algorithmic problems — arrays, sorting, intervals)
1 Ownership interview (behavioral)

Questions were not extremely hard, but you are expected to answer perfectly

Netflix

Process was long and slow:

• Recruiter screen (basic fit)
• Hiring manager (detailed discussion of skills and experience)
• Phone screen: product sense (metrics), ETL design (fact/dim), plus Python coding

Onsite:

• System & solution design (design pipelines and debug scenarios)
• Software & data structures (Python + Spark coding)
• Recruiter: culture fit
• Partnership & analytic mindset (non-technical)
• Technical bar-raiser / impact (non-technical)

The software & data-structures rounds at Netflix were more challenging: we discussed transforming a single-threaded (Scala) solution into Spark and then into a real-time streaming implementation, evaluating trade-offs of different approaches.

Apple

Apparently apple allows candidates to interview at the same time for different roles, so I had 3 processes in a 4 month interval.
Apple’s processes are very team dependent and the recruiters tend to not share much information regarding each step.

Software & Data Engineer team1
Hiring manager screening: in depth discussion about experiences/skills
Phone screen: Refactor some code
Onsite:

• Code comprhresion (You are given a codebase, you are asked to make some change)
• System design (design a data pipeline)
• Behavioural round
• Spark coding and discussion (spark coding and discussion regarding internals and optimization)

The general feeling is that the questions were not super complicated, you are not asked any leetcode questions, everyone seemed very chill and i had a positive experience

Software & Data Engineer team2
Hiring manager interview: discussion about experiences/skills
Phone screen: a java multithreading java
Onsite: nope, did not get here this time, i’ll need to study Completablefutures in Java :D

I guess this team was looking for a more software engineer with some data skills than a pure data engineer

Analytics Data Engineer
Hiring manager discussion and coding (30 minutes skills + 30 minutes python coding)
Another phone screen: spark coding + sql

Onsite:

• Onsite 1 Behavioural
• Onsite 2 Data pipeline design
• Onsite 3 Coding (python)
• Onsite 4 Data modeling + schema evolution

This team was looking for a "pure" data engineer where no skills outside python/sql/pipeline were asked.

Preparation

I have mixed feelings about preparing for data‑engineer interviews. You usually don’t need to grind LeetCode or master advanced algorithms, but the path isn’t clear and teams expect different things.

Core skills
Python writing clean, testable code; common libraries; basic algorithms and data structures
SQL joins, window functions, aggregations, CTEs, performance considerations and index/partitioning implications
Data modeling facts vs. dimensions, schema design, normalization vs. denormalization, schema evolution, Slowly Changing Dimensions

Secondary skills
Distributed systems & storage Spark fundamentals plus the broader ecosystem: object stores (S3), table formats (Iceberg, Delta Lake), ACID transaction concepts, columnar formats (Parquet/ORC), partitioning and compaction strategies
Streaming Kafka fundamentals, stream processing (Spark Structured Streaming, Flink), basic event time/watermark concepts, exactly‑once vs at‑least‑once semantics