American Express Data Analytics Interview Questions and Answers (2026 Guide)

Quick answer: American Express's data analytics hiring in India runs mainly through its Credit & Fraud Risk (CFR) Analytics & Data Science team in Gurugram and Bengaluru. Reported processes range from a lean 2-round path (resume and ML fundamentals, then a Director-level case study) to a 4-round path with a coding test, two technical rounds and a leadership round, and Glassdoor's own rating categories for Amex show real variation rather than one fixed process. SQL, core statistics, classic machine learning and fraud- or credit-risk-flavoured case studies recur across accounts.

About American Express

American Express is a global financial services company built around charge cards, credit cards and merchant payments, with about 76,800 employees worldwide as of end-2025. (Yahoo Finance) India is a major hub: American Express opened what it describes as its largest office worldwide in Gurugram in 2024, a nearly one-million-square-foot, LEED Gold-certified campus in Sector 74A. (Business Standard) Amex's own careers site describes its Analytics & Risk Management group as working with “one of the most valuable data sets in the world”, spanning Credit Risk Analytics and Fraud Prevention & Compliance. (American Express Careers) India data-role hiring is concentrated in the Credit & Fraud Risk (CFR) Analytics & Data Science Center of Excellence, based in Gurugram and Bengaluru, working on predictive models for fraud, credit risk and marketing decisions using Python, SQL, SAS, Spark and Hive. (American Express Careers, Analyst-Data Science, Gurugram)

American Express's Interview Process

Glassdoor actually runs separate rating categories for Amex depending on job title, and they do not agree with each other, which is itself useful signal:

• Data Analyst (Glassdoor aggregate): 3.2 out of 5 difficulty, 82% positive experience, average 3 days to hire across 11 submitted interviews. Reported topics: work experience, SQL, Python and statistics, plus logic puzzles, a business case study on identifying “high-value” customers from transaction data, and a behavioural round built around the STAR method. (Glassdoor)

• Data Scientist (Glassdoor aggregate): 3 out of 5 difficulty, only 53% positive experience, average 33 days to hire, described as commonly 4 rounds: recruiter screen, technical screen, and a case-study round testing statistics, machine learning and Python, R or SQL fluency. (Glassdoor)

• Analyst – Data Science (a separate Glassdoor job-title category, matching Amex's actual India posting title): also 3 out of 5 difficulty, but 92% positive experience and a faster average 16 days to hire. (Glassdoor)

• One detailed candidate account for a Gurgaon Data Scientist opening describes just 2 rounds: the first on resume projects, core ML concepts and guesstimates; the second with a Director, combining a case study on predictive modelling with a discussion of Amex's business model. (InterviewQuery)

• A separate off-campus account on GeeksforGeeks for a hybrid data-science and engineering role describes 4 rounds instead: a HackerEarth coding test (arrays and graphs), a technical round on univariate and bivariate data, wide versus long data formats and probability and normal-distribution questions, a second technical round on multithreading, HashMap versus Hashtable and a classic puzzle, and a leadership round with an Engineering Director on an algorithm problem and code walkthrough. (GeeksforGeeks)

What Candidates Report

Across these accounts, recurring technical themes include handling class imbalance in classification models, when to use recall over precision, causes and fixes for overfitting, and the difference between Random Forest and Gradient-Boosted Decision Trees. (InterviewQuery) Fraud-detection case studies and logical or probability puzzles are reported specifically for Gurgaon and Bengaluru interviews, which fits the CFR team's actual mandate. One prep-site summary frames the overall Amex data-science bar as resting on three things: SQL depth, familiarity with fraud-detection and financial-data patterns, and cultural fit against Amex's own “Blue Box Values”, a real, named set of company values. (American Express) That three-part framing is the aggregator's own synthesis, not a quoted candidate report, so treat it as plausible shorthand rather than a confirmed rubric.

What We Could Not Verify

Glassdoor's India and US pages for Amex returned access errors on direct fetch, so every Glassdoor figure above comes from search-indexed summaries rather than pages we read in full. We could not reconcile the 2-round Director-case-study account with the 4-round GeeksforGeeks account: they may reflect different teams, levels or years rather than one fixed pipeline, so we are not presenting either as the universal Amex process. We found no India-specific employee headcount for Amex; a 2025 Yahoo Finance report gives a global figure of 76,800 but does not break India out separately. AmbitionBox returned nothing directly attributable to a data-analyst or data-scientist candidate account in this research pass.

How to Prepare

• SQL depth, not just syntax: joins and window functions come up across every account we found, and one aggregator flags SQL as carrying particular weight.

• Core statistics and classic ML trade-offs: class imbalance, recall versus precision, overfitting, and Random Forest versus Gradient-Boosted Decision Trees are all specifically reported.

• A fraud- or credit-risk-flavoured case study: since CFR is Amex India's main data-role hiring track, practise reasoning about a fraud-detection or credit-decisioning scenario, not just a generic dataset.

• Logical and probability puzzles: reported specifically for Gurgaon and Bengaluru rounds, alongside the standard resume and project walkthrough.