The questions below represent the types of technical and behavioral inquiries you will face during your Amex GBT interviews. They are designed to test your practical knowledge and how you apply it to real-world scenarios. Use these to identify patterns in what the company values, rather than treating them as a strict memorization list.

Past Projects & Behavioral

These questions assess your hands-on experience, your ability to articulate your past work, and how you handle workplace challenges.

• Walk me through the architecture of the most complex data pipeline you have built.
• Tell me about a time you found a significant data discrepancy in production. How did you handle it?
• Describe a situation where you had to push back on a product manager's data request.
• How do you ensure your data pipelines are documented and maintainable by other engineers?
• Tell me about a time you optimized a slow-running process. What was the impact?

SQL & Data Modeling

These questions test your ability to query complex datasets and design structures for analytical reporting.

• Write a query to calculate the rolling 7-day average of flight bookings per corporate client.
• Explain the difference between a star schema and a snowflake schema. When would you use each?
• How would you design a data model to track user interactions on our corporate booking platform?
• What are window functions, and can you provide an example of when you would use RANK() versus DENSE_RANK()?
• How do you handle slowly changing dimensions (SCD) in a data warehouse?

Coding & Architecture

These questions evaluate your programming proficiency and your ability to design scalable data systems.

• Write a Python function to parse a directory of CSV files, filter out invalid records, and aggregate the results.
• How would you architect a pipeline to ingest real-time currency exchange rates and apply them to historical booking data?
• Explain how you would use Apache Airflow to manage dependencies between multiple daily batch jobs.
• What strategies do you use to test your data pipelines before deploying them to production?
• How do you handle out-of-memory errors when processing large datasets in Python or Spark?

This visual timeline breaks down the typical stages of the Amex GBT interview loop, from the initial recruiter screen to the final technical and behavioral rounds. Use this to pace your preparation, ensuring you are ready for the endurance required during the multi-interview stage. Keep in mind that specific rounds may vary slightly depending on the exact team and region.

Deep Dive into Evaluation Areas

Past Projects and Experience Deep Dive

Your past experience is a cornerstone of the Amex GBT evaluation process. Interviewers will dissect your resume to verify your hands-on experience and understand your depth of knowledge. Strong performance in this area means you can articulate the business value of your past projects, explain your specific technical contributions, and defend the architectural choices you made.

Be ready to go over:

• End-to-End Pipeline Design – How you ingested data, transformed it, and delivered it to end-users.
• Technical Trade-offs – Why you chose a specific database, framework, or cloud service over alternatives.
• Failure and Recovery – Instances where a pipeline failed in production and how you resolved it.
• Advanced concepts (less common) – Strategies for handling late-arriving data, schema evolution, and real-time streaming architectures.

Example questions or scenarios:

• "Walk me through the most complex data pipeline you built in your current role. What were the bottlenecks?"
• "Explain a time when you had to migrate data from a legacy system. How did you ensure data integrity?"
• "Describe a situation where your project requirements changed midway. How did you adapt your data model?"

Coding and Algorithms

As a Data Engineer, you must be proficient in writing clean, scalable code to manipulate data. This area is evaluated through practical coding exercises, typically focusing on Python and SQL. A strong candidate writes optimized, readable code and proactively considers edge cases, null values, and performance implications.

Be ready to go over:

• Advanced SQL – Window functions, complex joins, subqueries, and query optimization techniques.
• Data Manipulation in Python – Using libraries like Pandas or PySpark to clean, aggregate, and transform datasets.
• Core Algorithms – Basic data structures and algorithms, particularly those relevant to data parsing and transformation.
• Advanced concepts (less common) – Object-oriented programming principles applied to custom ETL frameworks.

Example questions or scenarios:

• "Write a SQL query to find the top three spending corporate clients per region over the last fiscal year."
• "Given a raw, nested JSON file of travel bookings, write a Python script to flatten and clean the data."
• "How would you optimize a query that is currently timing out on a table with 500 million rows?"

Data Architecture and Modeling

Understanding how to structure data for analytical consumption is critical. Interviewers will test your knowledge of data warehousing principles and your ability to design scalable architectures. Strong candidates can design logical and physical data models that balance storage costs, query performance, and ease of use for downstream analytics.

Be ready to go over:

• Data Modeling – Star and snowflake schemas, dimensional modeling, and fact vs. dimension tables.
• ETL vs. ELT – Understanding the differences and knowing when to apply each pattern.
• Cloud Infrastructure – Familiarity with cloud data warehouses (like Snowflake or Redshift) and scalable storage solutions.
• Advanced concepts (less common) – Data mesh principles, automated data governance, and implementing CI/CD for data pipelines.

Example questions or scenarios:

• "Design a data model for a new hotel booking feature. What facts and dimensions would you include?"
• "How would you architect a daily batch pipeline to process 10TB of transaction logs?"
• "Explain how you would implement data quality checks within an automated ELT process."