Loadingone moment…

All interview experiences

Data engineering interview · Big Tech & Enterprise Software

Meta

SQL is king. A 5-SQL + 5-Python phone bar, practical Python streaming, product-sense data modeling, and a decisive Ownership round.

Stages: 55 with reported questions
Difficulty: Not rated
Cloud: Internal stack (Presto/Hive/Spark
Evidence: 1 of 2sources are first-hand

01

What Meta does

And what its data engineers work with.

Business: Social media / advertising
Platform: Internal stack (Presto/Hive/Spark, Dataswarm) — not public cloud
Stack named in its interviews: SQL
Python
Presto
Hive
Data modeling
Product sense

02

The loop

Stages in the order candidates meet them.

01Recruiter
02Phone screen (5 SQL + 5 Python)
03Onsite (3 full-stack)
04Ownership round
05Offer

5 stages. 5 rounds have reported questions, described below.

03

What each round expects

The topics it covers and the questions candidates reported.

ROUND 01
Phone screen
Coderpad — pass = 3+ of each
High confidence
- SQL
- Python
2 reported questions. See them below
ROUND 02
SQL
High confidence
- SQL
- Modeling
3 reported questions. See them below
ROUND 03
Python
High confidence
- SQL
- Python
3 reported questions. See them below
ROUND 04
Data Modeling / Product Sense
High confidence
- Modeling
- Design
4 reported questions. See them below
ROUND 05
Ownership (Behavioral)
30 min standalone
High confidence
- Behavioral
2 reported questions. See them below

04

Reported interview questions

By round. Questions marked Expected are written from the company's standards.

Reported from accounts and guides, by round

Phone screenHigh confidence

SQLHigh confidence

PythonHigh confidence

Data Modeling / Product SenseHigh confidence

Ownership (Behavioral)High confidence

Expected written from Meta's standards

Nobody reported these exact questions at Meta. They are the questions a loop for each of its top topics is known to ask, so prepare for them as likely, not as confirmed.

Return the top 3 rows per group, and say what happens on ties.SQL
Find the latest row per key in a table that has duplicates.SQL
Design a star schema for orders, and state the grain first.Modeling
How do you keep history when a customer's address changes?Modeling
Group events into sessions with a 30-minute gap.Python
Count by key in a file bigger than memory.Python
Tell me about a pipeline you owned that failed in production.Behavioral
Describe a disagreement about a design, and how it ended.Behavioral
Design a daily pipeline from an app database to a warehouse that is safe to rerun.Design
How would you handle late and duplicate events?Design

05

Must know

Ranked by how much of the evidence points at each topic.

SQLReported2 reported questions across 3 rounds
Data modelingReported2 reported questions across 2 rounds
Python and codingReported2 reported questions across 2 rounds
Behavioral and communicationIn the prioritiesNamed in the prep priorities
Pipeline and system designReported2 reported questions across 1 round

06

What matters most

Do the first group before the second.

SQL: multi-schema analytical, %-of-users, NULL handling
Do first (4)
Practical Python: tumbling windows, string/dict manipulation
Data modeling + product sense (metrics, dashboards)
Ownership STAR stories with measurable outcomes
Distributed concepts
Then
Meta product knowledge

07

Rejection patterns

Why candidates were turned down, as reported.

Not clearing the 5+5 phone bar (need 3+ each)
Weak product sense
Vague Ownership stories without measurable impact
Behavioral is named in 7 of 50 company accounts.
NULL/edge-case mistakes in SQL
SQL is named in 37 of 50 company accounts.

08

Reported experiences

6 reported. Each opens and closes. Composites are labelled.

6 accounts

Data EngineerDate not stated2025Candidate account (2025)
The phone screen had five SQL and five Python questions. The onsite had three rounds that each mixed SQL, Python and data modeling or product questions, plus one round on ownership. Questions included the share of Messenger video calls, unique logins, ride requests in 15-minute windows, and product dashboards.
What to take from it Speed matters. Practise short SQL and Python questions against a clock.
Data EngineerDate not statedExponent / Glassdoor
The onsite was three one-hour rounds. Each round followed the same pattern: about ten minutes on product sense first, then technical questions.
What to take from it Practise talking about a product's numbers before you write any query.
Data EngineerDate not statedExponent / Glassdoor
The technical screen was 60 minutes: 30 minutes of SQL and 30 minutes of Python. Candidates say 30 minutes goes very fast, so practise with a timer.
What to take from it Practise under time. Fast and correct beats clever.
Data EngineerDate not statedExponent / Glassdoor
A product question turned into a live search for the cause of a drop in an Instagram number. Other candidates got Instagram and Marketplace examples.
What to take from it When a number drops, list the likely causes and say how you would check each one.
Questions named
- Root-cause a drop in an Instagram metric
Data EngineerDate not statedExponent / Glassdoor
Candidates had to turn their product answers into tables: dimension and fact tables, and when to store facts in a flat form.
What to take from it Be ready to go from a business question to a table design.
Questions named
- Design dimension and fact tables for a product scenario
Data EngineerDate not statedExponent / Glassdoor
There is a separate behavioral round. One guide says it tests initiative: how you influence peers, partners and stakeholders.
What to take from it Prepare a story where you led without being the boss.

Sources and how sure we are

Strong first-hand signal. SQL dominates; the Ownership round is a genuine dealbreaker.

Medium — bobbyhsu, "Meta Data Engineer Interview Experience 2025"
IGotAnOffer — Meta Data Engineer interview (guide, not a first-hand account)

09

Rehearse the Meta loop

Their reported questions, one at a time, against a stopwatch.

Phone screen · 1 of 14

5 SQL questions

Out loud, in order: clarify, build it, check it, explain it.

00:00

10

Your preparation plan

Pick your level and the weeks you have.

Built from what Meta asks, strongest first. For the wider basics at your level, follow the Mid level plan alongside it.

Week 1
SQL
Read: SQL chapters Drill: SQL practice Say it aloud: SQL questions Full plan: Ace the SQL round
Data modeling
Read: Data modeling chapters Drill: Modeling practice Say it aloud: Data modeling questions Full plan: Ace the modeling round
Python and coding
Read: Python chapters Drill: Python practice Say it aloud: Python questions Full plan: Ace the Python round
Behavioral and communication
Read: Behavioral storytelling Say it aloud: Behavioral questions
Pipeline and system design
Read: Data pipeline chapters Drill: Pipeline practice Say it aloud: Data pipeline questions Full plan: Ace the pipeline round
Week 2
Rehearse the Meta loop
Run the question deckThen repeat your weakest topic once more.