Walmart Data Engineer Interview Guide – Experienced

Round 1: Preliminary Round (Screening Round) – Telephonic Interview (45

minutes)

In this initial phase, the focus was on discussing my prior experiences, particularly related to
data engineering tools and platforms I’ve worked with. I was also asked to elaborate on
some of the core data concepts and my work with them.
Key Discussion Points:

ar
 Overview of Previous Projects: I discussed my involvement with tools like
Mixpanel, Kafka, ETL processes, Datahub, Spark, and Presto architecture.
 Data Modeling: Detailed insights on how I created a data model during

ek
experimentation and A/B testing.
 Why Walmart? I explained my motivation for applying to Walmart, citing their global
presence, innovative data practices, and impactful role in the retail industry.

ad
Round 2: Technical Interview 1 (Coding & DSA Round)
W
Overview:
The second round of the interview focused primarily on assessing core technical skills
m
relevant to data engineering, particularly in coding, data structures, algorithms, and large-
scale data processing frameworks. This round lasted for about 1 hour and 30 minutes and
was conducted by a senior data engineer. The discussion covered various domains,
ha

including coding proficiency, SQL expertise, big data technologies, cloud platforms, and key
engineering practices.
ub

Topics Covered:
1. Data Structures and Algorithms (DSA):
 Medium-level data structures, including arrays, stacks, linked lists, and trees.
Sh

 Problem-solving related to fundamental algorithms.

2. SQL:
 Complex SQL queries, particularly using advanced features such as window
©

functions.
 Writing efficient queries for scenarios involving large datasets.
3. Big Data Concepts:
 Understanding of distributed processing systems, particularly related to
Apache Spark and Hadoop.
 Architectural and operational aspects of these big data tools.

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

 4. Cloud Computing:
 Scenarios related to cloud platforms such as AWS.
 Cloud service management and optimization techniques.
5. Python Coding:
 Writing Python code for automation and data engineering tasks.
 Coding challenges with a focus on algorithmic problem-solving.
6. DevOps & SDLC:
 Basic knowledge of DevOps practices, continuous integration/deployment
(CI/CD) pipelines.

ar
 Understanding of the software development lifecycle (SDLC) and agile
methodology.

ek
Detailed Breakdown of Interview Questions:

ad
1. Data Structures and Algorithm Questions:
 Coin Change Problem: A typical dynamic programming question asking for
the minimum number of coins required to make a change for a given amount.
W
This problem tests the understanding of optimal substructure and overlapping
subproblems, which are central to dynamic programming.
 Partitioning a Linked List: This problem requires partitioning a linked list
m
based on a value, ensuring that all nodes with values less than a given value
x appear before those with values greater than or equal to x. It evaluates
knowledge of linked list manipulation and partitioning techniques.
ha

2. SQL Questions:
 Finding nth Highest Salary: Given a table of employees and departments,
the task is to find the nth highest salary within each department. This is
ub

typically solved using SQL window functions, such as DENSE_RANK(), which

assigns a rank to each row within a partition of the dataset. The candidate is
also expected to explain why they chose DENSE_RANK() over RANK()—with
the former allowing ties to be handled without skipping ranks.
Sh

 SQL Query Design: The task involved writing SQL queries to identify
employees with the highest salaries within each department. This challenges
both logical thinking and SQL proficiency, as the solution requires correctly
implementing window functions or alternative approaches.
©

3. Big Data and Spark Optimization:

 Airflow and Kubernetes: Questions related to how Airflow operates in a
Kubernetes environment, including the use of Pods and the interaction
between the Airflow scheduler, web server, and worker machines.
Understanding containerization and Kubernetes' role in managing distributed
workloads is essential for big data engineers working in the cloud.
 Optimizing Spark Jobs: When Spark jobs take longer than expected,
performance bottlenecks are common. The candidate is expected to suggest
methods for identifying these issues, such as analyzing Spark UI logs,
checking for skewed data, and tuning configurations for resource allocation.

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

  Cluster Resource Allocation: Given limited resources in a Spark cluster, the
interview explores how to optimize resource distribution. This could include
strategies for configuring executors, adjusting memory allocations, and
managing job priorities
 Spark Job Code: A coding task was presented, where the candidate was
asked to write a code snippet to upload Parquet files to an S3 bucket using
the boto3 Python library. This test evaluates knowledge of both Python and
cloud-specific APIs for interacting with data storage.
4. Airflow and Logging:
 Airflow Log Storage: The interviewer asked about how Airflow stores logs
and the role of its backend database. This tests an understanding of Airflow’s
architecture and how it manages operational logs in a distributed

ar
environment.
5. Cloud Computing and AWS:

ek
 AWS-Based Scenarios: Scenarios revolving around AWS services were
discussed, focusing on real-world applications and how cloud platforms can
be leveraged to solve big data problems efficiently. This includes
understanding various AWS services like S3, EC2, Lambda, and their

ad
interaction with big data tools like Spark.

Key Insights for Candidates: W

 Coding and Algorithms: Proficiency in algorithms and data structures is essential
for problem-solving in coding interviews. Candidates should practice common
problems related to dynamic programming, linked lists, trees, and sorting algorithms.
m
 SQL and Database Optimization: SQL queries should not only be accurate but also
optimized for performance, especially when working with large datasets. Candidates
should be familiar with advanced SQL features like window functions and be
ha

prepared to explain their choice of techniques.

 Big Data Tools and Cloud Platforms: Knowledge of tools like Spark, Hadoop,
Kubernetes, and Airflow is crucial. Understanding how these tools interact in
distributed systems and cloud environments is key to succeeding in interviews for
ub

data engineering roles.

 Practical Application: It’s important to be able to write code under pressure,
especially for cloud and big data tasks like working with AWS and performing
Sh

operations in Spark. Prepare for hands-on coding exercises that test both theoretical
understanding and practical implementation.
©

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

 Round 3: Technical Interview 2 (Data Modeling/System Design with Big Data
Concepts)

Overview:
The third technical round, which lasted approximately 1 hour and 45 minutes, was focused
on data modeling, system design, and big data concepts. The interview was conducted by a
Staff Data Engineer from Walmart. This round required the candidate to demonstrate in-
depth knowledge in system architecture, big data tools, Java, and advanced data
engineering concepts, with a focus on both theoretical understanding and practical coding
abilities.

ar
Topics Covered:

ek
1. System Design:
 Event-driven architectures and large-scale system design.

ad
 Specific focus on designing systems like Mixpanel.
 Detailed exploration of load balancing, request handling, and system
components.
2. Big Data & Spark:
W
 Coding tasks with Spark, focusing on data ingestion and transformation using
Delta Lake.
m

 Optimizations for Spark jobs, including skewed joins, broadcast joins, and
Spark's Catalyst Optimizer.
ha

3. Java & Advanced Java Concepts:

 Questions on multithreading, synchronization, garbage collection, and
serialization.
ub

 Deep dive into Java collections, including interfaces, maps, and linked lists.
4. ETL and Data Warehousing:
Sh

 Understanding of data warehouse concepts, schema design, and ETL best

practices.
 Key topics included Snowflake vs. Star schema, normalization, and Slowly
Changing Dimensions (SCD).
©

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

 Detailed Breakdown of Interview Questions:

1. System Design and Event-Driven Architecture:

 Designing Mixpanel: The candidate was tasked with designing the Mixpanel
system, an event-driven analytics platform. The candidate leveraged tools like
draw.io to illustrate how events are captured from different sources (Android,
Web, and iOS apps). This exercise assessed the candidate’s understanding
of event-driven architectures and the flow of data between various
components in a distributed system.
 Request Handling in Distributed Systems: The candidate was asked to
explain how a request would travel from a client (like opening a Presto URL in

ar
a browser) through various system layers, including DNS resolution, load
balancing, and routing through the Presto coordinator. This question tested
the candidate’s understanding of networking and how requests are handled in

ek
complex, distributed systems.
 Custom API with Spring Boot: A hands-on coding exercise where the
candidate was asked to write a simple service and controller class in Spring

ad
Boot, simulating the development of a REST API. This tested the candidate’s
knowledge of Java, Spring Boot, and their ability to implement API logic
effectively.

2. Spark Optimization and Big Data Concepts:

W
 Upsert in Delta Lake: The candidate was asked to write code to read data
from a Delta Lake stored in an S3 bucket and perform an upsert (update
m

existing records and insert new ones) based on a primary key. This task
involved using Spark DataFrames, emphasizing knowledge of data ingestion,
transformation, and Delta Lake's capabilities.
ha

 Spark Optimizations: The interview covered various Spark optimization

techniques, such as handling skewed joins, using broadcast joins for small
tables, leveraging the Catalyst Optimizer for query optimization, and
ub

understanding the differences between repartition() and coalesce() for

managing data partitions in Spark.
 Spark Tungsten & Catalyst Optimizer: The candidate was asked about
Spark’s Tungsten and Catalyst components, which are critical for optimizing
Sh

query execution. Tungsten manages memory and execution for Spark, while
Catalyst performs query optimization. The candidate needed to explain how
these technologies improve performance in distributed big data processing.
©

3. Java & Advanced Java Concepts:

 Garbage Collection in Java: The candidate was asked to write code to
manually invoke the garbage collection process using the Java GC thread.
This tested knowledge of Java memory management and understanding of
how garbage collection is handled in the JVM.

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

  Multithreading and Synchronization: The interviewer probed deeper into
Java multithreading concepts, asking the candidate to explain synchronization
and write code to manage synchronized threads. This task required
understanding thread safety and how to use synchronization mechanisms
(like synchronized keyword) to avoid concurrency issues such as race
conditions.
 Serialization vs Deserialization: The candidate was questioned on the
differences between serialization (converting objects to a byte stream) and
deserialization (converting byte streams back to objects), and how they are
used in distributed systems for transmitting objects over networks.
 The transient Keyword in Java: The candidate was asked to explain the use
of the transient keyword, which marks fields to be excluded from serialization.

ar
This concept is essential for optimizing data transmission and storage in
distributed systems.

ek
4. System Design and Synchronization:
 Semaphore in Java: The candidate was asked to explain and implement the

ad
concept of a semaphore in Java, which is used for controlling access to
shared resources in concurrent programming. They were tasked with
completing code for a semaphore, managing processes and ensuring
synchronization to avoid deadlocks.
W
 Deadlock Prevention: The interviewer tested the candidate’s understanding
of deadlock prevention techniques, asking how deadlocks occur in
multithreaded systems and how to prevent them, specifically using
semaphores and other synchronization mechanisms.
m

5. ETL & Data Warehousing:

ha

 Snowflake vs. Star Schema: The candidate was asked to explain the
ub

difference between the Snowflake and Star schema designs in data

warehousing. The Star schema involves a central fact table connected to
dimension tables, while the Snowflake schema normalizes these dimension
tables into multiple related tables.
Sh

 Data Warehouse Design: The interview shifted towards designing a data

warehouse from scratch, with the candidate expected to consider new
requirements and structure a solution that fits the needs of a modern data
architecture.
 Normalization & Slowly Changing Dimensions (SCD): The candidate was
©

asked to explain normalization and how to manage historical data in data

warehouses using Slowly Changing Dimensions (SCD) Type 2. SCD Type 2
tracks changes over time by creating multiple records for a dimension,
maintaining historical accuracy.
 Onboarding Delta Lake Catalog to Presto: The candidate was asked how
to onboard a Delta Lake catalog to Presto, highlighting the need for
integration between big data tools for efficient querying and analytics.

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

  Agile vs. Waterfall: The candidate was asked to explain why Agile is
preferred over the traditional Waterfall model. This question focused on the
iterative, flexible nature of Agile, which is well-suited for data engineering
projects where requirements can evolve over time. The candidate detailed the
Scrum framework, sprints, Jira boards, and the benefits of incremental
progress.

Key Insights for Candidates:

 System Design & Architecture: A solid understanding of system design principles,

particularly in the context of event-driven architectures and distributed systems, is

ar
essential. Familiarity with tools like Spring Boot and techniques for handling large-
scale data processing in Spark is critical.

ek
 Big Data & Spark Optimizations: Proficiency in Spark, including its optimization
techniques (e.g., skewed joins, broadcast joins, and the Catalyst optimizer), is crucial
for tackling performance issues in big data workflows.
 Java & Concurrency: A deep understanding of Java, especially regarding

ad
multithreading, synchronization, garbage collection, and serialization, is essential for
solving concurrency-related problems and optimizing memory management.
 ETL & Data Warehousing: Knowledge of data modeling concepts like Snowflake
W
and Star schemas, normalization, and Slowly Changing Dimensions (SCD) is key for
building scalable and efficient data warehouses.
 Agile Methodology: Understanding of Agile principles, particularly Scrum, is
necessary for managing projects in a fast-paced, iterative environment. Familiarity
m

with tools like Jira and understanding Agile’s flexibility in adapting to changes is
critical for success in modern engineering teams.
ha
ub
Sh
©

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

 Round 4: Techno-Managerial Interview (Managerial Round): 1 hour 10 minutes

1. Introduction & Skillset Overview:

 Introduction of expertise and technical skillset.
2. Data Modeling & ETL Design:
 Questions on Data Modeling, Databricks, Datahub, PySpark, and architecture design
(ETL Design).
 Explanation of the Mixpanel project and data model creation using delta tables.
 Detailed explanation of the data pipeline on Databricks for creating aggregated tables

ar
based on business requirements.
3. Contributions to Open-Source Projects:

ek
 Discussion on contributions to open-source projects, including Datahub and Spark
Lineage.
 Explanation of Spark jar creation with Spark listeners and the Spline package.

ad
4. Cost Optimization:
 Questions on cost optimization in cloud technologies:

impact on your organization?
W
Can you share an example of a project you worked on that had a significant

 How did you contribute to cost optimization initiatives while working with cloud
technologies?
m

 Could you describe a specific cost optimization strategy you implemented in

the cloud and its results?
ha

5. Databricks & Spark Monitoring:

 Questions on capturing event logs and user activities on Databricks, including cluster
creation and job execution.
ub

 Questions on Spark monitoring and performance management.

6. Agile Methodology (JIRA & Scrum):
Questions on managing multiple tasks using Agile methodology.
Sh


©

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

Round 5: Director Round (Behavioral & Technical Round): 45 minutes

1. Introduction & Experience Discussion:

 Introduction: Introduction of self and brief overview of professional background.
 Project Experience:
 Discussion on the Datahub Spark Lineage Project and role as Data Engineer
at Meesho.
2. Core Principles & Values:
 Questions on core principles or core values of Walmart and personal inspirations.

ar
3. Team Management & Leadership:
 Situation-based questions like:

ek
 Tell me about a time when you faced a challenging situation at work and how
you handled it.
Questions on team management and leadership qualities.

ad


4. Technical Expertise:
 Discussion based on Presto vs. Spark distributed architecture, Databricks, AWS,
W
Delta Lakes, and Data Governance.
 Specific questions included:
 What is Avro file format & what is its significance in delta tables?
m

 Difference between Presto vs. Spark underlying architecture.

 Can Presto work with Near Real-Time Data (Streaming Data Source)?
ha

 How did you develop the Datahub using Open Source Projects such as
Spline & Datahub?
 What do you think about Data uncertainty?
ub
Sh
©

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar

 Round 6: HR Round (General Discussion & Salary Discussion): 30 minutes

1. General Discussion:
 Questions about experience with Big Data projects, hobbies, and strengths &
weaknesses.
 Inquiry about family background, previous interview experiences, and life goals.
2. Final Questions:
 "Why should we hire you?"
 "What inspires you to join Walmart?"

ar
3. Salary Discussion:
 Discussion around salary and benefits.

ek
4. Outcome:
 Positive feedback from HR, resulting in selection for the position of Senior Data
Engineer (Data Engineer-3) at Walmart.

ad
Glassdoor Walmart Review –
W
https://www.glassdoor.co.in/Reviews/Walmart-Reviews-E715.htm
m

Walmart Careers –
https://careers.walmart.com/
ha

Subscribe to my YouTube Channel for Free Data Engineering Content –

ub

https://www.youtube.com/@shubhamwadekar27

Connect with me here –

Sh

https://bento.me/shubhamwadekar

Checkout more Interview Preparation Material on –

©

https://topmate.io/shubham_wadekar

For personal use only. Redistribution or resale is prohibited. © Shubham Wadekar